I have been using Waters Carbohydrate column (4um Nova-Pak sperical silica bonded with trifunctional amino propylsilane) with 75/25 ACN/water at 30C to quantitate ~1 g/L glucose. The sample matrix also contains 8 g/L sodium chloride. The RI detector chromatogram has a large Cl peak that interferes with the glucose peak. Changing the mobile phase seems to help in separating the Cl and glucose peaks, however the retention times are getting quite long (25 minutes). Does anyone have any experience separating NaCl and glucose?

Dear Anon,
You might want to check with Bio-Rad, thye have some excellent columns for your purpose. You might try an HPX-87P or a Fast Carbohydrate Analysis column. Either way give them a call 1-800-4biorad or online at discover.bio-rad.com. These guys have been a great help in the past.

Regards,
Mark

Since you apparently don't need to quantitate the chloride, why not just remove the chloride before injection using something like Alltech's Maxi-clean Ag (silver) cartrideges? They fit onto a luer tip syringe.

Sometime I analyze glucose and derivatives in SEC/GPC system. System (polyvinylalcoholl column, dimethylformamide+LiCl as eluent and RI detector) is dedicated in our laboratory to analyze other samples, but works well for glucose. Glucose Rt is 12 min; salts are eluted in time 15 up to 22 min. At flow rate 0,5 ml/min pressure is 73 bar, then it is possible to reduce analyze time by increasing flow rate.
There should be similar usefully GPC with water or buffer as eluent.

Biorad 87H column, but after some amount of samples it will probably become an 87 Na column. In the Na form the column will still do the carbohydrate separation but there can be a large stationary phase volume change upon changing cations. I don't remember which way it goes. If you are unlucky it will create a void. Might use sulfuric acid as an eluant which would keep resin in protonated form. Separation may work on just the guard column and save $$$.

My problem is similar to the initial post. I want to get rid of the chloride ions in the urine for the analysis of arsenic species by LC-ICP-MS in urine samples. Although Ag is a good precipitant for chloride, I really don’t want to introduce Ag into the ICP. Any suggestions?

What about all these examples of RP for sugars, that would get rid of the chloride. I have been considering RP for glucose... any good?

Retention is very low, unless you add a salt to your mobile phase. I used to do this analysis on Resolve C18.

Re: determination of arsenic species in the presence of a large excess of chloride ion, there are two solutions. In the first place, when using a silver form cartridge to remove chloride ion its often forgotten that all ions which have soluble silver salts will exit the cartridge with a silver counter ion. For some reason this is often described as a silver leachate but this is actually simple ion exchange. This can be corrected by placing a second hydrogen form cation exchange resin cartridge immediately after the silver form cartridge. This should quantitatively remove all silver from the effluent and prevent silver contamination of the sample.

The other option is to use a relatively high-capacity anion exchange column. A good example is the IonPac AS10 (available from Dionex) which is used for arsenic speciation in groundwater samples that often contain high levels of chloride. In this case, when separations are performed at relatively high pH the arsenic species elute well after chloride and present no interference problems.

Similarly, in the case of glucose above, this can be dealt with quite easily by separating glucose via anion exchange at high pH. Columns for this purpose (such as the CarboPac PA1) typically have fairly high-capacity and can tolerate samples with a fair amount of chlorides without significant interference. Of course, if you are utilizing electrochemical detection chloride won't interfere anyway as it is not detected under the conditions for detection of carbohydrates.

Chris

Re: remove Cl from urine. I have also been thinking along the line of Dionex column. The Ag/cation idea is certainly sound but may be expensive in the long term. Thanks for the suggestion anyway, will post again when successful.