[ikjadoon]

Hello! Hopefully a quick question on gel filtration column chromatography. So, small sample volumes (1-3% of the column volume) allow for better separation (according to all the gel filtration procedures I have seen). Here is some evidence:



I do not understand why, though. Why would you get bad separation with a large sample volume? Isn't gel filtration basically dependent on time (measured by column length and flow rate)? Why would adding more volume disturb that? It's not as if a large sample will move through any faster than a small sample (right?).

Someone on Yahoo Answers posted this answer, which I also, unfortunately, do not understand:

"in gel filtration the volume of eluent is what moves compounds in and out of the pores in the matrix.A large volume of sample means the sample components will not be introduced to the matrix as a tight "plug" and the sample will slowly spread into the matrix meaning the beginning and the ending of the sample will be far apart meaning the sample will be spread out over the gel matrix resulting in poor resolution of that component from any other"

What does he/she mean when they say "the sample components will not be introduced to the matrix as tight plug"? Does tight plug mean tight adsorption? Why would sample volume change adsorption? And don't we want "the beginning and the ending of the sample will be far apart"? Don't we want separation?

[Arkcon]

Try to envision a chromatography column as a stack of disks.  You want not so much a "plug" traveling down the column, but instead a disk of analyte, individual components interacting, some more and others less, so the disks of analyte components separate from each other and exit the column for detection.

Of course, nothing works that way, there is always a near Gaussian curve of components traveling down the column, especially in size exclusion (which should ideally have no chemical interaction between media and analyte.)  But still, we want to avoid "overloading" an analytical column.  Also, this is important, column loading amount is determined by the area of the column, not its volume.  Some people tend to forget that.

[fledarmus]

Up to a point, the only thing that you are doing by increasing the volume is changing the size of the peak. As Arkcon said, this is a near gaussian curve - when you expand it, you are increasing both the height and the breadth of the peak, but not changing the retention time of the peak. If you have two peaks separated by only a small amount, if your solution is dilute you may get baseline separation, but as you increase the size of the peak by increasing the sample volume, the peaks start to overlap with increasing breadth.

There is a point however, where you start changing the actual interactions between the components and the separation system. When the concentration gets too high, you can use up all the sites of interaction on your stationary phase, and remaining material just moves on past without being slowed down by interactions with the stationary phase. This spreads out the peaks and leads to tailing rather than gaussian shapes. At this point your separation efficiency really goes down.

[ikjadoon]

Thank you both, Arkcon and fledarmus, for the reply! Right, I do have some intuitive sense that the best separation would be a little disk of analyte moving through a huge column, but I can't put my finger on why that is better than if I had a huge "plug" of analyte going down the column. Oh, and I did not know it was truly measured from the column area; thank you for the correction! Just to check, do you mean the surface area of the column or the area of the "circle" in the cylinder (pi*r^2 where r is the radius of the column)?

Increasing the volume /or/ increasing the concentration [basically increasing the number of analyte molecules] = increase the height and width of the peak. I knew that it would increase the height of the peak, but now I see why it would also increase the width. It's chance how many pores molecule L will hit; some molecules of L will hit 10,000 pores, some may hit only 9,500 pores. While they [molecules of L hitting 10,000 pores and molecules of L hitting 9,500 pores] are the same molecule, they will have different retention times, thus causing a widening of the peak. The more molecules that I add, the higher the chance that some will fall outside the "average" retention time and cause wider peaks and overall decrease separation resolution.  This makes sense to me; or have I missed it completely?

And too many molecules being bad also makes sense: by adding more sample (volume or concentration), you are taking up the pores on the gel and thus decreasing the separation efficiency of the column.  This makes sense to me, too.

So, in conclusion, we want low numbers of molecules (whether that be through a small volume OR a small concentration). I guess, for whatever reason, they state low volume in gel filtration procedures for simplicity. Now, this all made sense until I saw this:



  So close, but so far away. Where have I messed up?