Hi. I'm using:

-Dionex IonPac AS12A (4x200mm) analytical column
-CD20 conductivity detector
-1.0 mM NaHCO/2.0 mM Na2CO3 eluent

I'm experiencing problems with chloride peak tailing/asymmetry. Attached is a chromatagram displaying this phenomenon. Sorry for the crude cut and paste job. As you can see the chloride peak tails off and even changes to positive slope with higher concentrations, creating a ghost peak. I am almost sure that this isn't a ghost peak from contamination in the system because different sources of Cl standard show the same result, and also blanks come out clean. Is there anyway to elimate this tailing effect? Does it arise from the carbonate eluent? Thanks for your help.

clpeak.doc clpeak.doc (48 k)

John,
Don't know the answer to your problem, but I really like the accompanying chromatogram. You communicated the situation very well.
I would not call it assymmetry or tailing. I see peak splitting, you may find troubleshoot aids helpful for causes of splitting.
Jim

Check with other peaks: it may be a bad column (all peaks should show similar shape) or contamination of your chloride standard.

It would have been nice to see the whole peak not just the baseline. This would give us a better idea of total shape.

In reply to some of your comments thus far:
-Not a bad column, new columns have been installed showing similar peak shape for chloride
-Not Cl standard contamination, as mentioned above multiple sources of standards have been used displaying same results, in addition field samples analyzed show same results
-Showing the whole peak: The rest of the peak is very symmetrical. If I displayed it at full scale (>100 uS)the tailing effect would be completely unobservable. However the tailing effect is great enough to be mistaken and identified as a nitrite peak (ret. time=4.34s) by the Peaknet software.

Thanks for all your input, keep it coming.

John,
Ivan's question means what is the peak shape of other analytes? For example, if Br- has a peak like the Cl- one with a little peak following it, you have a void volume somewhere, in your column or elsewhere in the system.
Jim

John- There are at least 3 candidates for the
appearance of the 2nd peak (1) column overload,
(2) the dissolving solvent is "stronger" than the
mobile phase or (3) there is a real component in
the injected sample. To me, it looks most like the
latter. Note that the resolution increases by
decreasing the concentration, which is exactly the
strategy for separating 2 closely eluting
components and also that this peak is linear w
concentration. To investigate (1), what’s the
linearity of the major peak like (in terms of
correlation coefficient)? Regarding (2) are the
samples being dissolved in mobile phase, and wrt
(3) a contaminant could be introduced during
sample prep (does the blank truly represent the
whole sample prep.?) for example, the contaminant
is a constituent of the dissolution matrix and
ensuing dilutions are made w mobile phase. If
preping the sample w mp is not an option at least
this hypothesis can be tested by improving the
resolution(changing the mobile phase or increasing
the column length).

John,

I am looking into your problem. In order to better understand the situation, though, it would be helpful to know more about the specifics of your operating conditions. In particular I'm interested in the following: flow rate, suppressor , suppressor operating mode, suppressor operating conditions, injection volume and details regarding your standard (how it was prepared and whether other analytes are also present). If you would prefer, you can e-mail me directly regarding this topic.