Sciencemadness Discussion Board - Grignard successes and failures

Grignard successes and failures

Magpie - 18-2-2016 at 16:17

I just finished preparing n-butylMgBr, a Grignard reagent, and wanted to post about the experience as it may prove helpful to some. There are so many threads about this important technology that I couldn't pick one. Hopefully people can express their successes and failures here and get some help if they need it.

ChemPlayer has posted a YouTube video showing that this reaction's fussiness may not live up to its reputation. But when I spend one day making butyl bromide and 3 days getting the bloody ether out of engine starting fluid I take all the precautions to prevent a failure.

For USA ether extractors I find that I like Johnsen's Premium starting fluid (50% ether) the best due to high yield: right at 100ml per 205g can. For Australians I have read that "Start You Bastard" is popular. (Sorry, I couldn't resist that.) I place this over 3A activated mole sieves overnight to dry. I likewise dry my halide that way.

Glassware is dried for ~1 hr at 110°C. This includes every piece that is used for ether and halide transfers.

A CaCl2 guard tube is placed at the top of the reflux condenser.

I keep my Mg turnings in an air tight jar and use them "as is." I obtained these in 2006 from Swany, a forum member as a trade. I've never had a failure using them.

A small piece of iodine (~20mg, I suppose) is added to the Mg turnings.

Today's Preparation
The following procedure is based on the preparation of n-hexyl alcohol found on p. 253 in Vogel's 3rd. I used a 1/3 scale.

A stirring bar (egg) was added to the 500mL RBF followed by 12.2g of Mg turnings. This was set up with a reflux condenser and a 250mL p-e funnel using a Claisen adapter. The RBF was placed in a bowl of warm water over a mag stirrer-hotplate. 53.6 ml of n-butylBr was mixed with 167ml of dried ether and added to the p-e funnel.

The iodine particle was added. Then ~20ml of the bromide/ether solution was added. The reactants were turning dark and I judged the reaction had started so I added 33ml of dried ether on top of the reactants per procedure.

Then the reaction didn't look like it was going so I heated the water bath up to ~45°C. Still no reaction. So I added another 20mL of bromide/ether. Still nothing. Being afraid of a runaway if I added too much reactant I held back. Eventually I began adding more reactant and the reaction then was clearly boiling. I removed the water bath and then added reactant as fast as I could without overwhelming the reflux condenser. Of course the reactants never got very hot as ether boils at 35°C.

Below are pictures of my setup and the product.

Comments, suggestions, and questions are welcomed.

Grignard in progress

finished product

blogfast25 - 18-2-2016 at 16:34

@magpie:

Bet you didn't oven dry the stirrer bar, huh?

So when are you going to use your butylating agent and on what substrate?

I'm by no means an expert on Grignards but always felt the paranoia about moisture to be a bit OTT. Sure, water destroys the agent but small amounts would only reduce yield a bit. Am I wrong?

[Edited on 19-2-2016 by blogfast25]

UC235 - 18-2-2016 at 16:49

Good quality Mg turnings are a bit hard to come by. I've had very good luck with material purchased from galliumsource. Just crush them a bit in a mortar and pestle beforehand for good luck.

As far as activating the Mg, I like to put a tiny crystal of iodine in the bottom of the flask before adding the Mg (dry). Heat the bottom of the flask gently with a free flame or heat gun until purple iodine vapors rise into the Mg. Allow to cool before adding ether/halide. My grignards always start immediately using that treatment.

I see you are using a short liebig. I prefer to use a high surface reflux condenser like a Friedrich if you have one. This can handle furious boiling from the ether and still condense it effectively.

Not having an effective press to make sodium wire or the desire to set up a ketyl still for small volumes, I've found NaPb to be highly effective. Just melt equimolar amounts of lead and sodium together in a covered metal crucible. They melt together with a moderate exotherm. You may need to hit the crucible with a hammer to get it out, but it forms a brittle intermetallic that is readily powdered in a mortar and pestle. With large pieces of sodium in ether the surface seems to passivate but that does not happen with this alloy and its huge surface area when ground up. Any residue of the alloy can be quenched with water without fear of fire/explosion.

I've seen a weird effect when I attempt to cut down on my ether usage. Since the grignard complexes ether, it actually raises the boiling point significantly. A ~3M EtMgBr solution boils at ~65C. Of course, the fumes still need to be cooled to 35C to condense. As the carbonyl is added (exothermic) the solution boils furiously as the b.p. falls. At any rate, I do not recommend using grignards this concentrated. It set into a mass of magnesium alkoxide before I could get all the acetone in. I still managed ~70% yield though.

[Edited on 19-2-2016 by UC235]

j_sum1 - 18-2-2016 at 16:55

Alas, "Start ya Bastard" is only 10-30% ether. There are probably other better sources for aussies.
https://jr.chemwatch.net/CWWS.RG.DEMO/MaterialService.svc/Ge...

Chemplayer has a video illustrating what happens when water is present in a grignard.
https://www.youtube.com/watch?v=h6QfJwaKz8E

gsd - 18-2-2016 at 17:28

Hi Magpie!

If the picture shows finished product then either you have used large excess of Mg or the reaction subsided prematurely.
Looks like the latter.

My take on Grignard is:

1) I always use freshly "turned" Mg. I have a bar of Mg metal which I take to nearby workshop and get small chips and turnings made as required.

2) Both DEE and THF are OTC in my country. So I always use THF. I also use Sodium metal to dry the THF.

3) To initiate the reaction, I use about 1 ml of EDBR (Ethylene Dibromide) which works much better than I2 crystal.

4) If I have to take more than one batch with a given halide the for all subsequent batches I use about 10 ml of product from previous batch to initiate the reaction.

gsd

Edit: lt is important to see that the halide used is neutal. My experience is even if small trace of mineral acid is present in it, it creates problem. If not neutral then shake it with a pinch of NaHCO3 and then use it.

gsd

[Edited on 19-2-2016 by gsd]

chemplayer.. - 18-2-2016 at 19:06

Good and important topic.

We've done a few Grignard reactions now and learned the hard way. Our magnesium turnings are not the best quality but we do as UC suggests and crush them up, then dry them inside the flask in the oven. The ether is reagent quality and dried by adding very thin sodium metal slivers; we use a decent carbon steel Japanese sashimi knife (!) to cut them and the surface area is probably better than you could get using wire.

Within these constraints, we consistently find it's the purity of our starting halide that is the limiting factor:

- Methyl iodide can be rough and ready and even just CaCl2 dried after washing (for a few weeks) and it works fine.

- Benzyl chloride needs to be super pure and dried (with some sodium carbonate as well to remove HCl), re-distilled, then carefully dried again, and even then it's very sluggish and requires an hour of reflux to really get going.

- Iodobenzene is somewhat intermediate - it seems to take a long time to start, but then works ok.

It could be that even slight traces of water in (some) alkyl halide compounds cause enough hydrogen halide to be generated via the hydrolysis equilibrium that this inhibits the reaction badly. This seems to be quite difficult to remove; even by regular distillation as the gas tends to come over and redistill. We found this especially hard to remove with benzyl chloride.

blogfast25 - 18-2-2016 at 19:48

Quote: Originally posted by chemplayer..

Within these constraints, we consistently find it's the purity of our starting halide that is the limiting factor:

- Methyl iodide can be rough and ready and even just CaCl2 dried after washing (for a few weeks) and it works fine.

- Benzyl chloride needs to be super pure and dried (with some sodium carbonate as well to remove HCl), re-distilled, then carefully dried again, and even then it's very sluggish and requires an hour of reflux to really get going.

- Iodobenzene is somewhat intermediate - it seems to take a long time to start, but then works ok.

Purity effects definitely notwithstanding, the comparison strikes me as a little skewed: even at high purity do we expect reaction rates to be identical when these three are such different animals? Consider for instance R-I bond polarity, due to different electron pushing effects by R and its effect on R-Mg-I formation rate. Not to mention you're comparing one chloride with two iodides...

[Edited on 19-2-2016 by blogfast25]

chemplayer.. - 18-2-2016 at 22:05

Sorry, it wasn't really meant to be a comparison. The point is that different compounds react very differently, and the ones that are slow in forming the Grignard seem to be much more sensitive to purity issues, especially the presence of acid.

Waffles SS - 18-2-2016 at 23:18

I have a good experience on Grignard reactions and i have to say there is no need for use high purity Mg metal or use anhydrous ether, you can easily make Grignard reagent even if you dont dry your glassware !

What will happen if your reagent contain small amount of water? How we can make Grignard reagent without anhydrous solvent or dry glassware?

In fact you need more energy for starting reaction between Mg and Alkyl halide for making Grignard reagent,if you provide this energy water will react with Mg and make Mg(OH)2 this will make your solution anhydrous and your glassware dry and your reaction will start.

You can provide this energy by Ultrasonic.

@Magpie, put your flask for a few seconds in Ultrasonic cleaning bath at 40-50c and you can see how easily reaction start.

I usually make Grignard reagent by Ultrasonic cleaner.

[Edited on 19-2-2016 by Waffles SS]

clearly_not_atara - 18-2-2016 at 23:49

"Mechanism of Grignard reagent formation. The surface nature of the reaction"

It seems that magnesium halides form on the surface, and it's possible that in the presence of water these halides might be less permeable to nonpolar molecules. MgI2 is mostly covalent, so it lets organic molecules through more easily; MgCl2 is more ionic.

MrMario - 19-2-2016 at 08:39

I always see people using Magnesium turnings, but could one also use Magnesium powder (Mg)?
What would be the advantage of using turnings vs powdered magnesium metal?

Magpie - 19-2-2016 at 09:12

Quote: Originally posted by blogfast25

@magpie:

So when are you going to use your butylating agent and on what substrate?

I'm by no means an expert on Grignards but always felt the paranoia about moisture to be a bit OTT. Sure, water destroys the agent but small amounts would only reduce yield a bit. Am I wrong?

I hope to soon make n-amyl alcohol using CH2O. Yes, I know I can buy it.

The main concern is the reaction not starting. As I said I spent a lot of resources preparing to make this reagent. I don't want a failure like happened to me and other students in a college lab, apparently due to wet ether.

Also, I don't like to unnecessarily lose yield. Due to its low molecular weight a little water goes a long way.

@gsd:

Yes, both the halide and the Mg were added at 0.5 mole. The Mg, however, was in very fine flake-like turnings and the residual weight may not amount to as much as it appears.

I had no idea there were so many individual experiences and opinions concerning Grignards. This is interesting.

gsd - 19-2-2016 at 09:23

Quote: Originally posted by MrMario

What would be the advantage of using turnings vs powdered magnesium metal?

IMHO following reasons:

1) Gram for gram turnings have much lower surface area than powder. For a metal like Magnesium which is prone to get attacked by atmospheric Oxygen and moisture, this is certainly an advantage.

2) In the reaction mass, "active sites" over Mg surface will be much lower in turning than in powder which will offer much better control over the reaction.

3) Making turnings is much easier than making powder.

4) During long storage, if turnings have lost their shine you can still use them by rubbing off the oxide layer with polish paper on few chips. If the powder is oxidised then their is very little you can do.

gsd

zed - 19-2-2016 at 15:38

Well, the few times I have run Grignards, I used granular Mg. Not really a powder. More like the texture of coarse U.S. Table salt.

Used warm Mg, a crystal of I2, and BenzylChloride in EthylEther. Drop the crystal of I2 from the top of the reflux condensor; it lands deadcenter at the peak of yer pile of Mg. Attach an addition funnel, let a few drops of Ether/Benzyl Chloride fall. Bingo!

Always started to sizzle, shortly after the first drop hit. Of course, I was using fresh commercial products. And, I realize some of you are improvising, with self-made reagents, or recycled Ether, wherein absolute dryness may be difficult to attain.

careysub - 19-2-2016 at 17:05

How pure does magnesium need to be?

How about using magnesium ribbon? A length of ribbon could be easily abraded to remove an oxide surface layer, then snipped up with scissors.

Rotometals sells magnesium ribbon in sealed pouches, and is described as "unoxidized" (mine is bright and shiny), and is 99.8% pure (they say).

I do see that "electron transfer between Grignard reagents and carbonyl compounds is extremely sensitive to the presence of trace amounts of transition metals in the magnesium" (Advances in Organometallic Chemistry):

https://books.google.com/books?id=05H9FYDPVvoC&pg=PA151&...

This:
https://www.researchgate.net/publication/244764942_Influence...
shows a 0.1% transition metal content reducing yield from 80% to 56%.

OTOH, according to this site:

http://devarchive.cnx.org/contents/1927489f-c820-4cc7-b1c4-d...

"For his original experiments Grignard used magnesium of a purity of 99.2%. However, it is now more typical to use 99.8% pure magnesium. It is important that the magnesium not be too pure since it is thought that the transition metal impurities catalyze the reaction."

Magnesium-aluminum alloys (1-8% aluminum) are widely used, but aluminum is not a transition metal.

[Edited on 20-2-2016 by careysub]

skip - 19-2-2016 at 19:42

Ever do one in a beer bottle ?

fluorescence - 21-2-2016 at 07:02

We prepared a Grignard some months ago in the lab to make Dicyclohexylphosphine. So we used fresh magnesium ground it half an our in a mortar to get the oxide layer away, added a small crystal of Iodine, everything in a Schlenk flask under dry Nitrogen atmosphere and heated it wih a lighter (never seen that you heat flasks with a normal lighter but for these purposes it works quite well) and then left it for 24h to form anhydrous Iodide, then the next day added absolutely dry ether (Schlenk) as well and some CCl₄ to even further catalyze the reaction but it wouldnt work. It boiled on its own and we left it for 6 hours then checked in the NMR and there was not a single peak of our product.

According to my supervisor in all the years where this had to be prepared by students it only worked out in 2 (!) experiments. So we changed it to a much more reactive and a bit more dangerous setup:

40 ml 2.5 M fresh n-Butyl Lithium in Hexane same preparation the syringe immediately caught fire the second I got it out of the Nitrogen enviroment. We tried it again, no result, the Chlorocyclohexane simply didn't react to the ltihiated compound.

Since we didn't know if there was any product and the Lithium Chloride formed in the reaction shouldn't be soluble in Hexane we got rid of most the solvent using a vacuum and a cooling trap. Problem was the Cyclohexylchloride didnt react at all and so we had accidentally concentrated the Buthyl Lithium to its maximum concentration.

So we fetched some tert. Butanol and I added a drop and if you know tert butanol you can boil potassium in that stuff and it won't react. Here if this unreactive alcohol was enough to cause a loud banging noise when I added it. I added some more and it splashed quite vigorously. Don't want to know what would happen if you added water to this.

So if anyone plans to do his grignard using Cyclohexylchloride forget it, it just doesn't react at all even with Alkyllithium compounds and also never mess with conc. Butyl Lithium compounds

[Edited on 21-2-2016 by fluorescence]

DJF90 - 21-2-2016 at 09:03

Quote: Originally posted by fluorescence

So if anyone plans to do his grignard using Cyclohexylchloride forget it, it just doesn't react at all even with Alkyllithium compounds and also never mess with conc. Butyl Lithium compounds

[Edited on 21-2-2016 by fluorescence]

You could have suspected that this would be the case based on the respective pKa values of the lithiated species (n-BuLi vs c-HexLi). Even though this isn't an equilibrium R-H + R'-Li <=> R-Li + R'-H, the pKa values give an idea as to the relative stability of the different alkyllithiums. As such, you can expect that n-BuLi is more stable than c-HexLi, and so the metallation could be expected not to proceed (as you observed). I reckon t-BuLi would do the trick nicely though (remember to use 2 eq. in order to drive the equilibrium by E2 elimination of the t-BuCl byproduct).

As for the Grignard formation, theres a whole bunch of reasons it might not have worked. I suspect the extended "activation" was part of the problem. I'd have just loaded up the magnesium on top of a crystal of iodine, heated from below to vaporise the iodine, then charge the ether and proceed without too much of a delay. If you still have troubles despite the usual tricks, its probably worth giving the ultrasonic bath a go.

EDIT: I found the preparation of c-HexMgCl in "Organomagnesium methods in organic synthesis" - find attached.

[Edited on 22-2-2016 by DJF90]

Attachment: c-HexMgCl.pdf (129kB)
This file has been downloaded 471 times

fluorescence - 22-2-2016 at 08:29

Thanks, but I already own every paper there is on that topic. I have details of all the things you can manipulate and change and it should work. I suspect, that the maybe Acros Organics sold a wrong compound. But doesn't matter its not my project and we already made something else. And for the BuLi. It was hard to find a paper that used BuLi so we ended up with a patent and they are usually not really usefull. The rest of literature suggested using elemental Lithium but my supervisor was against this. If I have time I will buy some Chlorocyclohexane for my home lab and test this out again but yeah as said it's not my work and I have quite much other stuff to do at the moment.

zed - 22-2-2016 at 14:33

If the literature suggests using elemental Lithium, try it.

In these matters, Lithium, Sodium, and Potassium........are not equivalent.

Also, of course, Patents are often full of lies.

DraconicAcid - 22-2-2016 at 15:03

I remember having to react a particular organometallic compound with a Grignard to repeat a literature prep, way back in my graduate career. Making the product was easy. Isolating it was not.

Delta - 23-2-2016 at 14:29

I read through the majority of this thread and skimmed the rest, and it looks like an important piece of information is lacking. Grignard reagents to my knowledge readily attack atmospheric CO2. fluorescence was the only person I saw to say they had performed the reaction under inert atmosphere. Your product may be contaminated with pentanoic acid.

Metacelsus - 23-2-2016 at 16:35

Usually, the solvent (ether) vapors act as a shield from the atmosphere. I've never had any trouble with carbon dioxide (but then again, I've only done the Grignard reaction twice, using bromobenzene).

zed - 23-2-2016 at 16:47

Yet, these reactions generally proceed well, without extraordinary measures to protect against atmospheric CO2.

Just as a Sohxlet full of LiAlH4, over refluxing Ether, is not ignited by air. Despite the fact that no inert gas is being used.

I assume that Ether vapor itself, acts as an inert atmosphere, and prevents very much CO2 from entering the reaction vessel.

Further, the constant loss of Ether vapor from the top of the reflux condensor, may prevent CO2 from entering at all, via constant flushing of the system.

I don't know about you guys, but one thing has been a consistent in my adventures with Ether; I always lose a portion of the volume during the course of a reaction. It like to escape.

Delta - 24-2-2016 at 17:27

Quote: Originally posted by zed

Yet, these reactions generally proceed well, without extraordinary measures to protect against atmospheric CO2.

Just as a Sohxlet full of LiAlH4, over refluxing Ether, is not ignited by air. Despite the fact that no inert gas is being used.

I assume that Ether vapor itself, acts as an inert atmosphere, and prevents very much CO2 from entering the reaction vessel.

Further, the constant loss of Ether vapor from the top of the reflux condensor, may prevent CO2 from entering at all, via constant flushing of the system.

I don't know about you guys, but one thing has been a consistent in my adventures with Ether; I always lose a portion of the volume during the course of a reaction. It like to escape.

Thank you for clearing that up. I have always used THF, and have had a few problems with carboxylic acid contaminants when not using an inert atmosphere. Maybe my reflux wasn't hard enough to exclude air.

Nicodem - 25-2-2016 at 07:51

Quote: Originally posted by fluorescence

40 ml 2.5 M fresh n-Butyl Lithium in Hexane same preparation the syringe immediately caught fire the second I got it out of the Nitrogen enviroment. We tried it again, no result, the Chlorocyclohexane simply didn't react to the ltihiated compound.

Since we didn't know if there was any product and the Lithium Chloride formed in the reaction shouldn't be soluble in Hexane we got rid of most the solvent using a vacuum and a cooling trap. Problem was the Cyclohexylchloride didnt react at all and so we had accidentally concentrated the Buthyl Lithium to its maximum concentration.

I never saw an example of Li/Cl exchange. As far as I know it does not work at all. You can even ortho-lithiate chlorobenzenes, after all. This reaction only works for Li/Br and Li/I exchanges (which is by the way among the fastest reaction ever observed, faster even than proton transfers). You would need to use bromocyclohexane with 2 equivalents of t-BuLi. To prepare alkyl lithiums from chloroalkanes you need to do it with metallic lithium just like for BuLi, though I don't see any need for it given that CyMgCl is commercially available.

chemrox - 25-2-2016 at 12:51

I've made a lot of phMgBr over the years and never had a problem starting the reaction. It's always good to start with a higher concentration of halide/ether mixture. I hear that THF is a little more dicey a solvent but haven't had to use it. The I2 trick usually overcomes any residual oxide. I have a shitload of Mg turnings btw if anyone needs some. The text book crap about putting the addition funnel on top of the condenser is dangerous and unnecessary. Simply reflux dry ether for a minute or two and no need for inert gas. Put in the first dose of more concentrated halide solution and go. A little heating and stand by with an ice water bath. These mothers run like crazy when started. I abhor the idea of using an LAH soxhlet; why would you do that zed?

CaptainPike - 25-2-2016 at 16:57

I can't add to the expertise of this thread (you guys are smart). Just have a couple of questions @Magpie.

What is the piece of glassware between the addition funnel and the side arm of the claisen? At first I thought it was just an adapter, but it looks like there is some liquid in it.

I don't imagine the water was turned on in your "reflux" condenser, prior to assembling the glassware, right?

Magpie - 25-2-2016 at 22:27

It's an adapter, 24/40 to 19/22.

Quote: Originally posted by CaptainPike

I don't imagine the water was turned on in your "reflux" condenser, prior to assembling the glassware, right?

The water can be turned on any time - the water is in a closed system.

MrMario - 26-2-2016 at 03:41

Maybe stupid question, but when is water allowed to get into the grignard again? I have never preformed one, so if one could maybe clear this up for me.

Magpie - 26-2-2016 at 09:23

R-MgX + H2O ---> RH + MgXOH

This not only is true for water but for any chemical with a sufficiently labile H, like a carboxylic acid. Even small amounts of water are harmful due to its low molecular wt.

Delta - 26-2-2016 at 09:35

Quote: Originally posted by MrMario

Maybe stupid question, but when is water allowed to get into the grignard again? I have never preformed one, so if one could maybe clear this up for me.

Water is allowed during workup, not during the reaction. Typically this is as a saturated NH4•Cl solution, or a dilute HCL solution. This is to protonate to the alcohol.

Delta - 26-2-2016 at 09:37

Quote: Originally posted by Magpie

R-MgX + H2O ---> RH + MgXOH

This not only is true for water but for any chemical with a sufficiently labile H, like a carboxylic acid. Even small amounts of water are harmful due to its low molecular wt.

Don't grignard reagents also react with the protons of alcohols and primary and secondary amines?

… yeah but

CaptainPike - 26-2-2016 at 13:20

Quote: Originally posted by Magpie

It's an adapter, 24/40 to 19/22.

Quote: Originally posted by CaptainPike

I don't imagine the water was turned on in your "reflux" condenser, prior to assembling the glassware, right?

The water can be turned on any time - the water is in a closed system.

I just meant that if a cold condenser were lying on your bench, it could conceivably condense water vapor from your lab area, which would run down into your reaction flask. Okay, so maybe I'm paranoid, but it could happen right? And maybe even miniscule amounts of water, added up…

Magpie - 26-2-2016 at 14:04

Yes, of course. I don't turn the water on prematurely for that reason. I just had no idea that is what you were getting at.

zed - 26-2-2016 at 17:53

"I've made a lot of phMgBr over the years and never had a problem starting the reaction. It's always good to start with a higher concentration of halide/ether mixture. I hear that THF is a little more dicey a solvent but haven't had to use it. The I2 trick usually overcomes any residual oxide. I have a shitload of Mg turnings btw if anyone needs some. The text book crap about putting the addition funnel on top of the condenser is dangerous and unnecessary. Simply reflux dry ether for a minute or two and no need for inert gas. Put in the first dose of more concentrated halide solution and go. A little heating and stand by with an ice water bath. These mothers run like crazy when started. I abhor the idea of using an LAH soxhlet; why would you do that zed?"

Best yields were reported for reduction via LiAlH4 ADDITION. Me, I hated the idea.