Chromatography Forum

I am looking at developing a method for detemining the amount of sulfite present in a formulation containing a tosylate salt. The tosylate salt to sulfite ratio is approx' 100:1. Conductivity detection has been attempted, but the large tosylate load affects the column; resolution and retention time decrease throughout the run. The use of SPE to remove the tosylate salt has proved unsuccessful as significant amount of sulfite is retained in the cartridge. I have a few ideas to try next, but I would appreciate any comments on the proposals:

1) Analyse samples by IC as per Dionex application note 54: ion exclusion chromatography with pulsed amperometric detection. Is the tosylate still likely to be an issue?

2) HILIC with ELSD. Again, would this eliminate the tosylate problem?

Thanks in advance.

My recommendation would be to use a porous graphitic column (such as Hypercarb) under one the conditions described in this paper:

Elfakir, C. ; Chaimbault, P. ; Dreux, M.
Determination of inorganic anions on porous graphitic carbon using evaporative light scattering detection. Use of carboxylic acids as electronic competitors
J. Chromatogr. A 1998, 829, 193-199.

The Tosylate salt is not one of the ions analyzed but the phenyl group will make the salt to be retained very strongly. Actually you might even have to use a combination of organic and high concentration of TFA to make it to elute, which would be perfect as you will have the minor compound first followed by the major. ELSD would be also perfect for this application, especially if you use a Sedere one which allows you to change the gain over the gradient time, (in case you are interested for the simultaneous quantitation of both ions)

mojo,

Sorry for the late response but I wouldn't use either method. I would use a column with low levels interaction with aromatics such a the IonPac AS20 from Dionex and simply dilute as necessary to obtain acceptable performance. With the tosylate at 100ppm the sulfite will still be 2-3 orders of magnitude above the detection limit with supressed conductivity so detection souldn't be an issue.

Chris