How much Water is there in our Methoxide

[dgs]

I've just posted on VOD some methoxide water results. To bring this post 'up to speed' on here the question I have been asking is not only how much water is in the methoxide but also where has the water come from.

When we dissolve KOH in methanol apparantly two different compounds are produced;

1 Potassium dissolved in methanol; this produces no water except for liberating some water in the potassium.

2 Potassium methoxide; this reaction produces water and as this water is produced it stops the methoxide making reaction. (320 gms of water per 1KG KOH would be liberated from this if the reaction went to completion)

I thought it would be interesting to see just how much water was in the KOH/methanol/methoxide.

According to the chemist 'Neutral' who did some work on this some years ago the methoxide making reaction comes to completion early on as the water it produces stops the reaction. Most of our 'methoxide' is therefore not methoxide at all but simply KOH dissolved in methanol.

Here are my results and an estimate of the completion of the methoxide reaction. There is an assumption that the missing 10% purity of KOH is water so I am the first to admit these results could be way out.



So I remembered to do a test on some KOH/methanol mix and got a result of 1.32% water.

I also tested the water in the methanol which was 450ppm.

OK, this is where it gets interesting, is there really 10% water in our KOH? Anyone ever tested it weigh/heat in 105deg oven for a while/weigh.

let's carry on as if there is 10% water in the KOH.

My concentration of KOH/Meth was 1kg KOH in 18 litres of methanol, so in 18 litres of methanol this would account for a water level of 0.55% from the water in the KOH.

The water in the methanol accounts for 0.05%

According to Countrypaul who gave me some figures a while ago; the amount of water produced from dissolving 1KG of KOH in methanol would be 320gms (this is assuming the potassium methoxide making reaction goes to completion.
In my case using 18 litres of methanol this would give us 1.78% water.

If we subtract the water in the methanol plus the water in the KOH from my result of 1.32% we are left with 0.72% water from the Potassium methoxide making reaction.

This means the potassium methoxide making reaction had reached 40% completion.

Thoughts Anyone

[Julian]

All good stuff but way beyond my knowledge of chemistry I'm afraid.  Will watch with interest though.

[dgs]

Probably beyond most of us Julian, good job Countrypaul is around, he has made some comments on the other channel.

I have made a slight mistake in that my water %'s are based on a W/V basis and should be W/W. On a W/W the water produced is 34% of the maximum amount possible due to the methoxide making reaction.

Paul did also confirm that there is a high probability of the missing 10% purity of KOH being water.

When I mix my next batch of 50/50 KOH/ASM 'methoxide' I will test for water as a comparison.

Could I do a water test on KOH using distilled white spirit as the solvent?

[dgs]

I intended to do a water test on the KOH/ASM methoxide I processed with yesterday and found I couldn't get the Bloody screw top off the Sandy Brae device. I'll have to use a couple of grips and test the next batch-------Bugga!

[dgs]

Because I keep forgetting to test the water content of my mixed KOH/ASM methoxide I mixed up some today to test. I will keep it for the next batch.

So the water content was 0.44% (compared to the 1.32% result from the KOH methoxide) Because the KOH amount is 50% of the original KOH methoxide I suppose a lower result was always predictable.

Because of the way that ASM works it leaves the methoxide water free, actually converting small amounts of water in the methanol. (please don"t ask me questions about this)  So I've always wondered if it could reduce the water content of mixed KOH methoxide. I suppose this result proves it does reduce that water, otherwise the result would have been around half the original.

Next thing I will do is test the water in ASM methoxide. (when I remember)

[dgs]

I am very suspect of my results on this thread.

It is accepted that any liquid containing methanol cannot be tested with the carbide manometer principal, simply because the chemical reaction produces water.

Apparently the Sandy Brae tester that uses Calcium Hydride does not undergo the same reaction (or that is what I was told)

The apparent rise in pressure after testing when the pressure has been released leads me to think water is somehow being produced between a reaction of the hydride and methoxide. Does anyone know. Countrypaul?

[Jamesrl]

IIRC country Paul wrote a post explaining exactly how much water is produced when mixing NaOH and KOH.

Might take a bit of searching I'm sure it's there somewhere. 

[dgs]

IIRC country Paul wrote a post explaining exactly how much water is produced when mixing NaOH and KOH.

Might take a bit of searching I'm sure it's there somewhere.

I think I have seen that Jim, but the trouble is the methoxide reaction is stopped by the formation of water and I don't think anyone knows exactly what level of water that is. As said after the test, probably the next day I released the pressure from the s/b vessel then re-tightened the top, only to find after several hours there was another build up of pressure. This happened 2 or 3 times, it can only mean more hydrogen gas was being produced from more water formation.

[countrypaul]

Dissolving NaOH into Methanol will give 18g of water for each 40g of NaOH - assuming 100% pure and 100% conversion.
With KOH the figures are 18g of water from 56g of KOH - same assumptions.

With most reactions the process is never 100% complete but sits somewhere in the middle, in the case of say aqueus calcium hydroxide and sulphuric acid this process is very much towards the production of calcium sulphate (insoluble precipitate), where as with aqueous sodium hydroxide and sulphuirc acid it is very much in the middle as  everything separates into the ions which all remain in solution.

With KOH (or NaOH) dissolving into methanol the reaction can be viewed as:

KOH + MeOH <--> KOMe +H2O

How much will be KOH vs how much is KOMe will depend on a number of factors, such as concentration, temperature, any contaminents, possibly even the material the container is made of.

If one of the compounds is removed, for eaxmple KOMe by transesterification of veg oil or KOH by making soap with the veg oil, then that would disturb the equilbrium so moving it more to one direction, as a result all other things being equal the equilibrium will be reestablished by converting more KOH to KOMe or KOMe to KOH.

This also means that if any attempt to measure the water level is performed during the reaction (transesterification as far as we are concerned), it may appear to increase water level as more KOH is converted to KOMe because the KOMe is used up. Problems could lie in the fact that glycerol absorbs water which could adversely impact any test results as this may or may not be reflected in the results.

[countrypaul]

I am very suspect of my results on this thread.

It is accepted that any liquid containing methanol cannot be tested with the carbide manometer principal, simply because the chemical reaction produces water.

Apparently the Sandy Brae tester that uses Calcium Hydride does not undergo the same reaction (or that is what I was told)

The apparent rise in pressure after testing when the pressure has been released leads me to think water is somehow being produced between a reaction of the hydride and methoxide. Does anyone know. Countrypaul?

I discussed this with Dave when I visited him a short while back, but never posted our thoughts at the time. So for the benefit of everyone else:

The Sandy Brae appeartus is normally made from anodised aluminium with a viton seal and probably some brass fittings (though we did not check those). One problem is that annodised aluminium (depending on the variant) can still be susceptible to corrosion by strong alkalis, and sodium methoxide is a very strong alkali. Whilst the reaction of aluminum with strong alkali is not that fast, it does release hydrogen - so the problem with the test may be due to the interaction of the Sandy Brae apparatus with the methoxide. This could be tested by trying the experiment without the calcium hydride and see if there is still a pressure increase over time. Problem is that this corrosion of the apparatus is not good and does damage the apparatus. Depending on the brass used, this can also corrode in alkali releasing hydrogen.

[dgs]

You could be correct Paul, as I said in my last post I am suspect of the results myself but not for the same reason as you.

I remember IMB on Infopop saying the Carbide manometer cannot measure water in Methanol because there is a reaction between the calcium carbide and methanol that actually produces water. Is there not a similar reaction between calcium hydride and methanol. When we discussed this I think you said you wouldn't be surprised if that was the case.

When I have tested just methanol in the past I have sometimes had the anomaly where the pressure still increases long after the reaction should have stopped, but that doesn't always happen eg certain batches of virgin methanol that have showed a result of around 300ppm.

It would be nice to find out what is going on here. I could get the Karl Fischer working again but it needs new solvent and titrant which from Sigma Aldrich is at least £150. I don't do enough water tests to really justify the cost. In any case it really needs to be kept in a controlled environment, temp and humidity wise not kept in a dampish outbuilding.

[countrypaul]

Calcium hydride could react with methanol to release hydrogen quite easily, not sure how I overlooked that. I would have thought it would be quite a quick reaction though - certainly quick enough to invalidate any figure for ppm of water.

Did your SB come with a table of temperature corrections, they are usually calibrated to either 20C or 25C and need the pressure to be corrected if used at a different temperature?

[dgs]

Calcium hydride could react with methanol to release hydrogen quite easily, not sure how I overlooked that. I would have thought it would be quite a quick reaction though - certainly quick enough to invalidate any figure for ppm of water.

Did your SB come with a table of temperature corrections, they are usually calibrated to either 20C or 25C and need the pressure to be corrected if used at a different temperature?

Indeed, I keep a careful check on the temperature when testing.

[Tony]

Is it worth adding a small amount of glycerol to the methanol before adding the KOH in order to absorb produced water?  (I'm assuming glycerol still forms a fraction by gravity in excess methanol?)

[dgs]

Is it worth adding a small amount of glycerol to the methanol before adding the KOH in order to absorb produced water?  (I'm assuming glycerol still forms a fraction by gravity in excess methanol?)

I think glycerol and methanol are miscible in all proportions.

I'm presently pm ing Countrypaul who asked me to give him details of the Karl Fischer titrant and solvent. In fact I've just pm'd him and lost the lot as I wasn't logged in. The forum keeps logging me out, anyone know why?

Just for interest sake I will do it on this thread as I'm typing anyway.

Instrument is a Karl Fischer volumetric KD 758 titrano.

Solvent is made by Fluka, hydranal solvent, contains imidazole, sulphur dioxide and methanol.

Titrant is titrant 5 made by Riedel de Haen. Contains methanol and Iodine.