Thursday, November 23, 2006

Anhydrous solvents

Most organic chemists need dry solvents from time to time and almost every single day you bump into someone who's looking for dry DMF, ether, acetonitrile, THF etc. When I was working in Cambridge this was never a problem. They simply have a still for every single solvent you could imagine. Above there's a picture from the still room in Cambridge. Pretty nice innit? A massive fire hazard but quite handy as long as it doesn't blow up. Now I've moved on and where I work now we've just been inspected by OH&S (Occupational Health & Safety) and they don't like stills because of the fire hazard etc. and I have to say I completely agree with them. Nice as they may be they are dangerous and more or less completely redundant. The solvents you can buy nowadays are of super high quality and do not require distilling so it's basically only distilled in an effort to dry it. Now I will concede that there are stabilisers in many solvents, in particular ethers, but it is very rarely something that will affect your chemistry. So what should you do? The perfect setup that will provide guaranteed anhydrous solvents every day consists of good quality super activated molecular sieves (MS) and a Karl Fischer (KF) apparatus. MS are expensive but you can reduce the cost significantly by buying bulk quantities. We used to get ours from Grace Davison in big drums and they were very very good. The Karl Fischer apparatus (see picture below) can be purchased from Metrohm. All you have to do when you have this set up is add some 3 or 4 Å MS (depending on the solvent) to your solvent close the flask tightly and when you come back the next day take a small quantity out with a syringe and needle and squirt it into you KF apparatus. The display will now show you how many ppm's of water there was in the volume you just added. Everyone doing anhydrous chemistry should have this set up. It's safe, you are always confident about whether your solvent is dry or not and it requires close to no maintenance. Unfortunately, molecular sieves will not dry everybody's favourite solvent THF so you have to hold on to one still. Also you cannot add anything with acidic protons to a KF apparatus successfully. It will think it's all water. So for example acetone or methanol wouldn't work. Regarding MS I believe that you should never attempt to dry them yourself and never recycle them unless they are going into exactly the same solvent (just bin them when they are dead). Many chemists think they are saving money when they reactivate MS. However, I seriously doubt that is the case with the amount of energy not to mention time required to do so. So in other words buy good ones and bin them when they stop working. If you aren't sure whether the MS you've got are any good put a couple in the palm of your hand and add one drop of water. If they get really really hot they a very good and if they only warm up a little bit they are rubbish. If you are in a situation where you can't get good MS I guess you will have to dry them. Apparently one way of doing this is to throw them in a microwave and nuke them on max power until they start glowing. At this point you have to stop immediately unless you want the entire microwave to melt and transfer the MS to a desiccator that you stick on a high vacuum pump (I haven't tried this myself so no guarantees). If you don't feel like burning your department to the ground there is always the good old vacuum oven that most departments have. Just heat them under vacuum for a couple of days and then stick them in a desiccator attached to a high vacuum line. Drying solvents overnight using good 4 Å MS should get acetonitrile, DMF, DMSO, dichloromethane, toluene, ether, 1,2-dichloroethane, chloroform and hexane down to a water content of 10 ppm or less. If you check your freshly distilled THF on the KF it should be around 15 ppm. So for successful living convince your boss or department to get one of these babies and shut down all those damn stills for good. D!

34 comments:

Anonymous said...

I suppose it depends on what you need your solvents for. Distillation also removes oxygen, which can be a big problem. Some solvents fail to be dried properly, even after distillation. For example, 1,2-dimethoxyethane requires several distillations before it's actually dry. Personally, though, I agree...for 99% of applications, activated 4A MS are enough. Frankly, I don't even flame dry my glassware for some of my reactions, so super dry solvents would be a total waste of time. I mean, there's no need to dry a silylation or peptide coupling, or even an HWE. Anyways, I activate mine at 350 C in a kiln overnight.

Anonymous said...

In the academic world, the fashionable alternative to stills are alumina towers. They also do a pretty decent job of deoxygenating solvents. We used at KF to monitor the dryness and it was competitive with traditional stills: THF from Na/benzophenone was 4 ppm, THF from column was 6 to 12 ppm, depending on the week we tested it.

Anonymous said...

Yeah, our group uses the Grubbs systems too. They work great, and they're way safer. The only drawback is their cost.

Anonymous said...

We got a Grubbs system during my PhD -- four towers, for Et2O, THF, CH2Cl2 and toluene.

For other solvents the boss used to recommend sucking the solvent through a column of alumina that had been baked overnight and cooled under dry argon. The solvent would go right from the alumina into the reaction flask. This is basically the poor man's dry still. It seemed to work OK but I never tested it with a KF apparatus. Any comments?

Anonymous said...

We have the same system. The Grubss system is pretty much activated alumina plus a copper catalyst for the solvents that can tolerate it. Testing with a KF apparatus is nice, but unnecessary in most cases...what kind of anhydrous applications do you want anyways?

Daniel Sejer said...

I would have to agree with all of you that the ultimate solution to dry solvents is The Grubbs system. It's safe, convenient, requires hardly any maintenance and the solvents are super dry. However, as someone pointed out this is an expensive solution and most academic institutions don't have the money to get a setup like this going. Also I agree that KF testing your solvents is unnecessary in most cases. 95% of the time what I test is perfectly dry. However, its nice to confirm that the solvent is dry. When my reactions don't work I like to know that wet solvents weren't the problem. D!

Anonymous said...

I want to get a feel for just how dangerous stills are. Personally, I know of two instances of still related fires and I have been working in labs for 10 years. The first was a girl didn't turn the tap off to the collection vessel properly and was distilling ether with an IR lamp. The ether dripped onto the lamp and whooosh up she went. The second was a DMSO distillation under vacuum where the thermometer was held in by a rubber bung (stupid). The girl applied the vacuum, the thermometer held for a little while but gave in eventually and whoosh another fire. Anyway, that is 2 in 10 years and department wide as these stories travel so it encompasses a lot of people doing a lot of distillations. So are stills really that dangerous when compared to other incidents, I know of 6 lab fires/explosions off hand that had nothing to do with stills.

I see the point though of the columns and would dearly love to replace my stills with them. They are, with regard to my stories above.....fool proof, convenient and if they stop even 2 fires every 10 years it is worth it.

What are others experiences with still fires per year, and please don't spare the detail, the mind boggles at what some people do. I once had a student put a beaker of ice in the oven in an attempt to make 'dry' ice.

Daniel Sejer said...

The still room at Cambridge University was introduced because there were many accidents involving personal stills. How many accidents they had prior to abolishing personal stills I don't know. However, even if stills don't blow that often they do blow occasionally. One of my former colleagues had a good friend that got taken out by an exploding THF still. He survived but doesn't look to nice after his severe burns. Molecular sieves do not explode so I think that is a much safer and simpler alternative for many solvents. Moreover, there is the issue of quenching stills. I know of several sodium fires as a result. Including one I was attempting to clean up myself. D!

Anonymous said...

In the 4.5 years it took me to do my Ph.D. we had 3 fires with our THF still alone. Two happened because someone forgot to close the stopcock on the arm used for withdrawing solvent with a needle. At least one of these instances was the result of someone sneaking into our lab via a back door and taking solvent when no one was around. One fire happened when I was quenching the same THF still (it was a 5 L beast). We never had any problems with our other stills (ether, toluene, acetonitrile, methylene chloride, and triethylamine). However, the lab directly above us had their ether still spontaneously explode for reasons unknown. It shook our lab and a guy was nearly rocked off his stool in the lab directly above the explosion. On our floor of the building, another group had quenched their THF still and left it in an empty fume hood to just sit overnight, in case anything happened. At 7 a.m. something did happen. It blew. Fortunately someone was in the lab to clena up the mess but not in front of the hood at the time of the "festivities". As I wrote, all of these events happened in a 4.5 year span. My take on this is that alkali metal stills are bad news. Calcium hydride and KOH are no problem at all and can be trusted. My suggestion is to use alumnia towers for THF and diethyl ether (and other ethers, for that matter) but traditional stills for everything else should be fine. ANd for the record, alumina towers are not that expensive. I have been told by a prof at a Canadian school that is looking at having a departmental solvent purification room that you can get an alumina system for 5 solvents for $30,000 (Canadian, about 27K US) for everything included (pump, tower heater for activating the alumina, solvent cans and all the piping and tubing). My neighbour here had a quote of $27,000 for 6 solvents but that did not include the pump and some other things. In the grand scheme of things, that is not a lot of money so, no offence folks, I don't buy the expense arguement. If you are in the academic world, talk to your chair and your dean and spin it as a safety issue because, well, it is!

Anonymous said...

There are of course safety issues with solvent stills, like there are with many other things in the lab. Luckily I haven't had any problems so far and I would like to ask all the other readers how many percent of the explosions/fires/any other problems were caused because someone was just doing the wrong thing? I haven't heard of many problems when the instructions were followed.

Anonymous said...

Stills blow up all the time. I'd say the major causes of accidents are: failure to turn the cooling water on, letting the still heat a very low level of solvent, and improper quenching. I don't call 27K "cheap"...it may well be worth it, but that's a year's salary for a grad student! Also, don't forget the cost of periodic column replacement, argon, and maintenance. Our lab has ether, THF, toluene, and dichloromethane on columns, and everything else as a still. Our dichloromethane seems to dissolve some of the alumina and jam up the dispenser a fair bit. So I'd say it works well, but it's not perfect, and certainly not as cheap as a still. You could probably get away with 1K per solvent for a still.

Anonymous said...

D.S. or anyone else?

Why would adding water to molecular seives cause them to heat up? Obviously some kind of exothermic rxn, but what?? Thanks.

Daniel Sejer said...

I've been thinking about why molecular sieves get warm when you add water for a long time. I don't have a clue. I suspect that if I had paid better attention when I had physical chemistry as an undergrad I might have been able to suggest something... I would like to know why this happens so if anyone can offer an explanation that would be cool. According to the MSDS for molecular sieves they can heat water to the boiling point upon addition!
Regarding alumina towers I wouldn't go as far as calling them cheap. Moreover, there's the whole issue of where to put them? I assume that they would have to go in a hood. Hood space is valuable and often there isn't enough of it. Also there's the whole maintenance issue. I don't know how expensive these things are to run. Do you have to buy special solvent drums for these things or can you just pour stuff from any old Winchester through them?
And finally a question: Do alumina towers remove stabilisers etc. from solvents? If not this could be a problem for some people where these interfere with there chemistry. D!

Anonymous said...

Solvent systems do not have to go into hoods, although they should be put in well ventilated areas. In principle, the solvent should never see the air, so a hood shouldn't be required. In principle. Ours sits in an equipment room, firmly outside a hood.

I don't think the upkeep is that high. We buy our solvents in drums (it's much cheaper), and without stabilizer, which can mess with the columns after a while. We transfer the solvent periodically into the solvent kegs under argon. The kegs take a lot of solvent, and can be weighed with a known tare weight, so refilling only has to be done every few months.

I don't know a thing about the heating sieves...

Anonymous said...

I recently got a quote for a 7 column system and it was AUD 50,000 which is about the same cost as a basic HPLC set up or GC set up. As for the ongoing costs of a system, it is the same for stills eg electricity, argon, solvent etc and the columns last something like 100L and a cheap option is to have the canisters repacked. There is also less solvent waste with a column system. I like the saftey angle, that is what we're playing.

Unknown said...

It's certainly better than the industry alternative of buying in dry solvent. Sure, they're dry for the first few uses, but that SureSeal doesn't last long... Chemists notice this, and use a fresh (or close to) bottle nearly every time. Result? Loads of half-empty bottles...

Anonymous said...

It's quite simple why molecular seives heat up when you add water. Its just thermodynamics. Its actually the opposite of why you feel cool when water evaporates off your skin. The seives bind water, hence lowering the kinetic energy of the water molecules which in turn gives off heat.

BlueBaron said...

We have an Alumina Tower system that has gone through about 100 L of THF without the need to replace it. It is also one of the best ways to dry DMF. Maintenece is easy, and we use unstabilized solvents.

Those DrySolv bottles are never dry enough. I got some DME from them and it never worked, but distilling my own and holding it over sieves and sealed inside a dessicator has left it good. I recommend just using teflon caps and quickly swapping the cap for a septum flowing with Ar (like a balloon) for each use. The septa degrade all the time and you get water and floaties in your stuff.

Preparing about a Liter of solvents or reagents by setting up a temporary still solves the problem if incompetent neglect that usually causes these explosions. It is only around for a day or two.

Anonymous said...

In our lab we have 3 systems for dry solvents. We have a traditional dry still for toluene. For THF we have the grubbs system. For DCM we just purchase predistilled dry kegs. Of all 3 systems I like the kegs the best and the still the least. The reason the grubbs system isnt my favorite is because of the vacuum system used to pump the solvents out. If someone is not careful they can pull a good amount of solvents into your vacuum pump.

Anonymous said...

Is it possible to titrate cyclohexanone using a Karl Fischer apparatus?? - I have heard that it contains a lot of water.....

Daniel Sejer said...

KF titration on anything with reasonably acidic protons, including ketones, isn't possible. I found this out the hard way when I decided to determine the water contents in a bottle of acetone. The technician that had to regenerate the KF titrator wasn't impressed. In general always add one drop of whatever you are titrating first to see what happens. If it goes nuts don't add more. I also discovered that benzaldehyde can't be used with a KF apparatus. I think that you can easily solve your cyclohexanone problem with some 4A molecular sieves. I am confident that some nicly active sieves wil dry your cyclohexanone to crispness. D!

Anonymous said...

You can perform Karl Fischer titration in solvents with acidic protons if you use the proper KF reagents. There are reagents available from EMD and probably other sources that are specifically for ketones. We use the ketone reagents to titrate water content in glacial acetic acid and it works fine.

Daniel Sejer said...

KF titration is the oxidation of SO2 using I2 and an appropriate base. I guess that employing a weak base in the KF titration could make the titration of "acidic" substances (e.g. acetone) possible. Thanks for the tip. That could become very useful down the track. There is a post on KF titration at wikipedia that may be of interest to some readers (http://en.wikipedia.org/wiki/Karl_Fischer_titration). D!

Unknown said...

We have two of these solvent stills...ACN and THF. We do some very sensitive spectroscopy and noticed problems with the ACN coming out of the still. It's causing a reaction to take place with our compounds. Has anyone else ever had this problem? Any suggestions?

Daniel Sejer said...

Heather, when you say still I assume you mean a solvent purification system (The things where the solvents run through a column and you just open a tap at the bottom to collect it)?
I don't have any experience with these things yet. Maybe someone else can help? However, if all you need is anhydrous acetonitrile I would recommend just adding some activated 4 Å molecular sieves (as described in the post). This normally gets the water contents of acetonitrile below 10 ppm overnight. D!

Unknown said...

Hi Daniel, Yes, I did mean solvent purification system. I need my CH3CN as clean as I can get it, and we use it all the time, so we set this up so that we would have a constant clean source. But, it appears that this is not the case. We definitely seem to have something in the solvent this reacts in a catalytic way with our compounds. This does not happen when we use the CH3CN out of the bottle, and I guess I was just curious if anyone had any idea why?
Thanks!

Anonymous said...

Can anyone refer a company that can repair/recharge solvent systems for THF, DCM? Our DCM flow is really poor and the column has not been serviced in the last five years. It's about time it got repacked.

Daniel Sejer said...

Nadia, I'm not sure how many people still read the older posts so I'm not sure how much feedback you'll get on your question. My best suggestion would be to contact the manufacturer of your system for a service. DCM has a poor reputation because the columns tend to clog up so if it's been 5 years the service is long overdue. If for some reason you cannot contact the systems manufacturer I can recommend the company glass contour: http://www.glasscontoursolventsystems.com/. D!

Anonymous said...

Why would adding water to molecular seives cause them to heat up?

http://en.wikipedia.org/wiki/Enthalpy_change_of_solution

Unknown said...

A remark if you want to dry ACN: don´t use 4A MS as stands in the post. The ACN molecules will get into the MS (they are small enough) so they won't capture H2O anymore.
Use 3A MS instead!

For the rest nice post man, thanx!

Anonymous said...

molecular sieves should have some kind of color indicator to show the water content :)

Anonymous said...

Can anybody comment on the paper JOC 2010, 75, 8351–8354?
It shows that MS are more than acceptable for THF drying while distillation (Na,benzophenone) showed quite poor performance.

Bryn said...

In answer to the query about Na/Bzp not drying THF.

Na is not the ideal alkali metal for drying THF because the boiling point of THF is 66oC and the melting point of Na is 98oC. The "Na" in the system is in fact NaOH supported on Na and hence there is no contact between the alkali metal and the THF. The correct choice of desiccant is K, with a melting point of 63oC. This has an extremely high contact in the molten state and water will be essentially exhausted in less than a minute.

Also note "blue" isn't dry, but only deoxygenated. As the still approaches dryness the blue will deepen to purple and finally crimson as the bzp ketyl goes all the way to the dianion. Oxygen reacts catalytically with the ketyl and water stoiciometrically.

If lower boiling solvents (such as diethyl ether or pentane) must be thoroughly dried then Na/K alloy must be employed, which is a slightly more hazardous drying agent. The best discussion on the subject is the late Peter Plesch’s book here - http://books.google.be/books?id=EXFeF62iL64C&lpg=PP1&dq=inauthor%3A%22P.%20H.%20Plesch%22&pg=PA116#v=onepage&q&f=false

One should also note that almost all still fires are ethers (usually THF). This is due to the buildup of peroxides to dangerous levels because the still has been left too long. A THF still is perfectly safe as long as it’s quenched frequently. It’s the ones that run months and months with only adding in fresh solvent and desiccant that go boom.

Wildfyr said...

I'm still not sure why anyone thinks that 3A sieves are no good for drying THF. Aside from mentioning it, Curly arrow doesn't provide any actual proof. On the other hand https://doi.org/10.1021/jo101589h shows that 38-72 hrs over 3A sieves produces sub-10 ppm water content. Na/Benzophenone left it at 45 ppm (wet was 108). Good enough for me.