Choose the pore size based on the sizes (molecular weight) of the molecules that you want to keep inside and the molecules you want to let out. These pore sizes are not exact, and the shape of the molecule can be a factor. Leave plenty of leeway to prevent loss of the molecules through the pores, but not so much as to sacrifice the speed of the process. For example, if you have a 10,000 Da protein and you want to dialyze it to get rid of a low molecular weight salt like NaCl, a 3,000 MWCO (molecular weight cutoff) membrane would be a reasonable choice. You could use a 1,000 MWCO membrane as well, but then the dialysis would be much slower.
Thank you so much for your great and valuable answers. For protein samples, we can select a membrane based on the molecular weight of protein and cut off value of the dialysis tube. How can I choose them for the purification/separation of carbogenic quantum dots? Is it possible? The size of the quantum dot is
I would guess that quantum dots have a considerably higher density than proteins, so the molecular weight of proteins would not be a good guide to the MWCO to chose for dialysis membrane.
According to theWikipedia entry on dialysis tubing:
"Pore sizes typically range from ~10-100 Angstroms for 1K to 50K MWCO membranes." That is 1-10 nm, so you should choose a membrane with a very low MWCO, like 1000 or 2000 Da if you want the quantum dots to stay inside.
For carbogenic quantum dots, you can roughly purify it first through some membrane filter. After that do the mass spectrometry of the sample. This will give you a rough idea about the molecular weight of qds. Then you can choose the dialysis membrane of your interest.
It can be assumed that the membrane shows pores that hold molecules above a given size and that are permeable to small molecules. A membrane should be selected with a Molecular Weight Cut-Off (MWCO) between the MW of molecules to be recovered in the sample container, and the MW of molecules to be extracted (usually salts, reaction by-products, ...). The link given below may be of help to you.
That would not be my preferred method. I don't think it would work very well, even if you could get dialysis membrane with such a low molecular weight cutoff. Size exclusion chromatography is a better method. You might also be able to separate the molecules based on other properties besides size, such as hydrophobicity or charge, using chromatography.