[rolnor]

I am working with a reaction that is sensitiv to the prescence of water.
I am using silylating agents like HMDS and BSA to remove water but it is possible that they somehow takes part in the reaction and I want to prove if they are.
I suspect this because BSA gives higher yield then HMDS and both should remove water completely. I can not use electrophiles like acid chlorides or silylating agents like TMSCl.
I am going to try N,O-bis(trimethylsilyl)trifluoroacetamide as this is more reactive then BSA.
I have tried 4Å molecular sieves but that is significantly less effective, the yields drop.
I also am thinking about calcium hydride, that would be completely different but maby less efficient. I can not use sodium/benzophenone or other reducing agents.
One important thing is that ROH is formed in the reaction and the role of the silylating agent can be to transform this to RO-SiMe3, that would explain why molecular sieves does not work, but again, HMDS should do this almost as good as BSA so there is a question whether the silylating agents have another role in the reaction.
I want suggestions on drying agents, the reaction is taking place in DMF or toluene.
Thanx

[wildfyr]

Trimethylorthoformate reacts with water to give methanol, which can be evaporated off.

[rolnor]

Yes, but only under acidic conditions and I can not have acid present.

[Babcock_Hall]

Do you mean water in the solvent, water that is produced in a reaction, or something else (atmospheric water, for example)?  Burfield and Smithers saw about equal effectiveness for 3 angstrom versus 4 angstrom sieves in drying DMF, but had very good results with sequentially using 3 Å sieves.  I am not sure that this is relevant to your application.

[rolnor]

I mean water in the solvent and probably also to remove ROH that is formed in the reaction. It is a bit unclear but this is the theory I am exploring.
Silylating agents work well and they do both. Calcium hydride would remove water forming CaOH2 and also ROH forming CaOR2 I suppose?

[wildfyr]

I have a ton of faith in sieves. How are you doing it with them? Are the fully activated? Fully activating them requires some fairly considerable asskicking conditions. 250°C+ under vacuum for a few hours is preferable.

Are you leaving them in the solvent long enough? Different solvents have quite different drying kinetics with sieves. 3A is preferred over 4A. You should also load it at 10-20% m/v.

[clarkstill]

I have a ton of faith in sieves. How are you doing it with them? Are the fully activated? Fully activating them requires some fairly considerable asskicking conditions. 250°C+ for a few hours is preferable.

Are you leaving them in the solvent long enough? Different solvents have quite different drying kinetics with sieves. 3A is preferred over 4A. You should also load it at 10-20% m/v.

Agreed. It's been posted many times I'm sure, but for the effectiveness of sieves in solvent drying, see:

https://pubs.acs.org/doi/full/10.1021/jo101589h?src=recsys

[rolnor]

I use 4Å, they are activated at 250°C under vacuum. I use 3g sieves in 15ml DMF-reaction solution, the DMF is stored over 4Å sieves also. It would be nice to test water-content with Karl Fischer but I dont have that in the lab. The sieves get very hot on contact with water so they are activated. The reaction is runt at 124°C, I filter of the sieves before heating under N2. The sieves would not remove the ROH formed and for now that seems as a good explanation why they dont work.

[phth]

I use 4Å, they are activated at 250°C under vacuum. I use 3g sieves in 15ml DMF-reaction solution, the DMF is stored over 4Å sieves also. It would be nice to test water-content with Karl Fischer but I dont have that in the lab. The sieves get very hot on contact with water so they are activated. The reaction is runt at 124°C, I filter of the sieves before heating under N2. The sieves would not remove the ROH formed and for now that seems as a good explanation why they dont work.

I would increasing the amount of drying power by upping to something like ground 3Å MS 90% w/v. Did you try to use toluene and a dean stark trap? Can you obtain a molecular sieve with a larger radius like 5Å? One could put the dessicant in top of the DS trap to remove water and alcohols...

[rolnor]

The alcohol molecule formed is much too large to be trapped by molecular sieves unfortunately.
I think the amount water present is much to small to be removed by DS-trap at least if I use it the conventional way but I will try that.
To ground the sieves is an interesting option, I have done that when doing Sharpless epoxidation with good result.