Chromatography Forum

Hello

I am working on quantification of galacturonic acid with this column. The samples also contains some monosaccharides : glucose, rhamnose, arabinose, xylose, galactose, mannose, fucose.

The detector is a ELSD. I got only 6 peaks in various conditions (flowrate, temperature) with these 8 compounds.

Is it due to the kind of column (maybe it has not a good selectivity for monsaccharides) or has somebody an idea to separate each compound?

Thank you for your answer.

All of those are available as pure compounds. First step would be to run standards individually and see what coelutes (or doesn't elute, as the case may be).

The mannose, galactose and xylose are coeluted

The Resex ROA is a sulfonated polystyrene packing in the hydrogen form. The same basic packing is also available in Ca, Ag, and Pb forms, each of which give different selectivity for carbohydrates (it's been a while since I worked with them, and I don't remember which one would be specifically applicable). Check with BioRad or with Dionex; they should be able to make a specific recommendation (web site info is in the FAQ; drill down to the "additional information" section).

I am no expert in carbohydrate analysis but from the little I know it will be impossible to separate all compounds on a single column (I am referring to the classic ligand exchange approach).
Switching, lets say to a lead form will resolve mannose, galactose and xylose but arabinose, fucose and mannose will most likely co-elute.
You can check out:
http://chromatography.transgenomic.com/ ... tChart.asp

That is a pretty useful chart. If you have more questions I can give you the contact info of the person in charge there. My e-mail is:
rhaefele@hamiltoncompany.com

Maybe someone from Dionex can shed some light on the capabilities of anion exchange in NaOH with a PAD...

cheers,

Robert Haefele
(NOT affiliated with TBIO)

Here is another idea: since the columns with Ca or Pb give a completely different selectivity, you could use a second column to resolve the ones that are unresolved on your H column. You will need to work out, what is best - just two different columns in series, or a column switching protocol, where the unresolved peaks from the H column are parked on say a Pb column and eluted from that after the analysis form the H column is complete.

It's a bit more complicated, but if you like to tinker, this could work.

Here is another idea: since the columns with Ca or Pb give a completely different selectivity, you could use a second column to resolve the ones that are unresolved on your H column. You will need to work out, what is best - just two different columns in series, or a column switching protocol, where the unresolved peaks from the H column are parked on say a Pb column and eluted from that after the analysis form the H column is complete.

It's a bit more complicated, but if you like to tinker, this could work.

i agree. add another column in series. either a rezex RCM or RPM would give the best separation. using only the ROA is not enough to separate all of the compounds.