[Nix]

Hi there,

I'm new to TLC and am attempting to do some semi-quantitative measurement of a component in an adjuvant that was diluted. As a standard, twice I have run plates with three spots composed of 6, 12, and 24 uL of a solution of known concentration. These were made by adding 6 uL once, 6 uL twice, and 8 uL thrice per spot, allowing for time to dry between spots so as to keep the spot size small.

On two different plates I have gotten the same result, with the dots appearing after vanillin staining but in a curve, despite what appears to be an even solvent front. (See below) Can anyone explain this and what I might do to counter it? My solvent is chloroform:methanol:water at 90:10:1.

Thanks for any advice!

[Babcock_Hall]

If the plate touches the filter paper in the chamber, one sometimes sees artifacts.  Can we rule this out?

[Nix]

I think so, yes. This result has been reproduced three times now, but the last one was better (less curved). Not entirely sure what could be different, except that I added a bit of fresh running buffer.

[MOTOBALL]

I would suggest that you do two things.

1. Measure 10 cm up the plate from the origin and scrape away about 0.5 cm of silica gel right across the plate, so that the solvent can only run 10 cm for each spot.

2. Run two plates. First has light , medium , heavy loading as usual.
The second has light, heavy, medium loading.

The results should detect whether edge effects are in play or whether it is just a question of sample overload.
If sample overload, a thicker layer of silica gel may be necessary.

Please show us the results.

[MOTOBALL]

Looking at your mobile phase, I would not be surprised to see edge-effects in play here.
Chloroform and methanol will both evaporate much faster than water. In the center of the plate, the mobile phase is surrounded by mob phase on all sides, and will stay in equilibrium at 90:10:1. At the edge of plate, the mob phase has an empty space adjacent. The equilibrium will not be maintained. You may actually have a mob phase of 90:10:x, where x = 2, 5, 10, 20 etc.
Looking closely at the two plates, even the light loading seems to be slightly in advance of the central spot.

[wildfyr]

This is probably a result of how the plate is placed in the liquid phase. If the corners touch first, then solvent will quickly wick up the sides and give a bell shaped curve appearence to the solvent front, and it takes time for it to even out. I suggest you use a shallower amount of solvent in the bottom, and spot your sample a little further from the bottom of the plate. Another way to fight this it to use a wider TLC plate and avoid spotting near the ends.