Tlc

[liambob1]

So for tlc, in selecting the correct mobile phase and stationary phase what are the factors which influence what ones you choose, how do you decide what ones to use?

[Horza2002]

That all depends on what you are TLCing. If you have a very non-polar molecule (basically only containing carbon, hydrogen and halogens), then a non-polar solvent system will be needed (e.g. hexane, hexane/diethyl ether, petroleum ether). If you have a slightly more polar molecule (containing protected nitrogens, oxygens, sulphurs or lots of them) then you'll need something more polar (hexane/ethyl acetate, ethyl acetate, dichloromethane, acetone). If you have an even more polar specicies (carboxylic acids, unprotected nitrogens, phosphonium salts) then you'll need something very polar (DCM/methanol). You can't really go above 10% Methanol though as you start to affect the homogenousity of the silica gel.

 

Those are the basic rules, but normally you have to play around with varying system until you find one that works. I normally start at 50% ethyl acetate in hexane as a baseline and then change in from there. It does massively depend on the compound that you are using though.

 

O, and the stationary phase is also important. Normally, silica gel is acid, and so amines do not move as they get protonated. There are several ways around this problem; either add a second base (normally triethyl amine) that neutralises the acid silica so your amine doesnt stick. Alternatively you could use alumina (aluminium oxide) as the stationary phase, which is basic in nature.

Edited March 1, 2011 by Horza2002

[liambob1]

We did this lab today but it didn't work out very well. The sample we had to analyze was sucrose, so we dissolved it in ethanol and used ethyl acetate as the mobile phase and the tlc plate with silica as the stationary phase, the solvent front went up but no spots were detected.

 

In the second part we dissolved and l-aspartic acid the sucrose in water and used 15 mls of ethanol and 10 mls as the mobile phase and the tlc plate with silica as the stationary phase, but that didnt work either!

 

What went wrong?

Edited March 2, 2011 by liambob1

[Horza2002]

What where you using to visulise the spots? Im guessing that you used permanganate solution...

 

It could be that you didn;t put enough sample onto the TLC for you to be able to see any spots forming...it also be that the solvent systems were too polar and you compound just raced off the top of the plate. However, if you didn't see any spots at all anywhere in the plate, I would say that you either didn't put enough sample on or you used the wrong visulising method to detect them.

 

If you used straight ethanol on the second one, then you would have serious problems of the silica dissolving and therefore giving you heterogenious stationary phase to flow through.

[liambob1]

The plates were treated so that they would floresce under uv light and so the spot where the sample was would not floresce and thus we could see the spots. Also the solvent front was advancing very very slowly in the second plate, the assistant said this is why water should never be used as part of the mobile phase, so maybe that effected it. Also is it permissible to use the same solvent you disolved the sample in as your mobile phase?

[Horza2002]

...I have never heard of that visulisation technique before...how did you do it? Or was it you just help the plate under a UV light? If thats the case, then the compound needs to have some functional groups active to UV light (e.g. aromatic rings and double bonds). In the case of sucrose, I would be very very suprised if was visible with UV light...you should have used permanganate or vanillin as the stain...sucrose would show up VERY brightly using those stains.

 

You shouldn't use water as a cosolvent no...the water will be able to pick up protons from the silica gel and so you will have localised differences in the pH which will therefore alter your retention time. Some solvent systems naturally just run slower than others...

 

While yes, you can use the same solvent for the mobile phase that you used to dissolve the sample in; however I wouldn't recommend it. Typically you should use a solvent that dissolve your sample extremely well and so if you use it as your mobile phase, it will move very fast as far up the plate. So it yes indeed possible, but not advisable.

[liambob1]

I think the idea was the plate fluoresed and the sucrose didn't therefore the spots were the sucrose was were identifible because they didn't fluorese while the rest of the plate did

[Horza2002]

What did you do to the plate before you placed it under the UV light? As I said though, a permanganate or vaniliin stain would have been much muhc better tand showing surcrose

[liambob1]

We buy in the plates pretreated to fluoresce. Thanks

[Horza2002]

Yer those pretreated plates are not really that great. We have them and you can hardly ever see it. Its far far far FAR better to use a TLC stain; have a look at one of my posts in organic chemistry for a list of stains and what they do.

[icepeaks635]

for tlc u need to choose the right solvent and the right stationary phase like silica gel.

solvent are like methanol ,ethanol,

Edited March 8, 2011 by icepeaks635

[Dan_Ny]

Yeah, like Horza2002 said, you need to take some stain and the usual silica tlc plates. It is the most simple thing on earth. I don't know the exact properties of sugars on silica, but try mixtures of methanol and DCM or ethyl acetate/DCM or ethyl acetate/hexanes as eluents (mobile phase) or something in that polarity range.

 

A good way to find a good "mobile phase" (with the desired retention factor) for your compound is just to make 5 different solvent mixtures at once, prepare 5 different tlc plates at once, put them in there, run them and stain them (spares time). Take the best solvent mixture or change polarity again after that.

 

The sugars themselves should not be visible in the UV (if you use the nomal tlcs, that is), but KMnO4 should definitely stain your sugars in high enough concentrations. Otherwise, have a look at Horza2002's list of stains...

Edited March 9, 2011 by Dan_iel