New to IC, and having some issues with older instrumentation

[Akurth]

Background:
I am new to IC, and working for a small environmental company. I was left with 2 Dionex-120 systems, and no real instruction on how to maintain or troubleshoot these systems. The work being done is simple anions analysis (F, CL, NO2, Br, NO3, PO4, SO4) in the standard matrices, using carb/bicarb eluent.

Equipment: Dionex -120/AS40 autosampler, AERS 500 4mm Supressor (Brand New), Dionex IonPac AS-14/AG 14 250/4mm

Issue: I have started seeing reductions in concentrations on QC samples. The reductions started with PO4, which is getting a ~70% return. The Calibrations are also failing for PO4. Within the past week, I have seen reductions in almost all analytes, with 80% recoveries on QC samples. The difference is the calibration curves for the other 6 analytes are excellent. I do run soil samples on one of the 2 instruments. I have followed all the cleaning procedures outlined in the manuals for the Dionex column, and have almost rebuilt both systems over the past 6 months. My retention times have shifted, but not drastically, and not out of my 10% window used for RT studies. The columns are not new, and have had quite a few samples run on them, but nothing out of the ordinary based on past run logs.

I am wondering what can be done to bring my recoveries back, and what can be done to prevent this from happening in the future.

I appreciate any response, this can be frustrating for a person with a minimal chemistry background.

[postb8822]

Poor recovery on a single ion out off a group of several would lead me to rule out the LC in many ways. If you had issues with the autosampler, or detection, you would expect to see similar behavior on all ions since it's just conductivity. I am assuming you are injecting standards that contain all ions at once, and not individual solutions?

If you are injecting individual ions then all options are on the table. Re-reading your post, now I see that you are starting to see recoveries slip on multiple ions. Are you using a suppressor? When you say calibrations are good or bad, what are your criteria for accepting those?

[M_Farooq]

When you start seeing losses with troublesome ions like phosphate, chances are that your column and / or suppressor is contaminated with metallic impurities. I wonder if your soil samples are rich in aluminium and iron, since both of them form precipitates with phosphate. Note that the suppressor could be dirty as well (with these metallic ions).

Did you give acid washes to both suppressor and the column? Since you have two instruments, the practical approach is
1. Use the same column with phosphate standard on both instruments. See the difference in peak heights in conductivity. Do the peak heights or areas compare well or not? If not, the suppressor is the culprit in one of the instruments.

2. If you see similar peak heights for phosphate on both instruments, it could be the contaminated column itself.

[Akurth]

Thank you for the responses. The suppressors on both instruments are less than 2-3 weeks old and have not had soils run on them. The system with the very low phosphate recovery (80%) is the one that soils are normally run on, so the suggestion of the column having metal contamination could be spot on.

The failing recovery of chloride and sulfate are the biggest questions to me. Both will have an r2 value of >.999 when the curve is run, then an internal calibration verification following the calibration will see both with recoveries of 85%. I can try an acid flush again, but I am worried I am going to start to degrade the column.

[FenderChrome]

Sorry if I'm a little late to the party. o-Phosphate can be tricky and as others have pointed out it is likely due to contamination of the column with metals. One trick that sometimes works, is simply changing out the frits in the columns (assuming your using dionex columns with removable frits). I would start with the frit on the inlet side of the guard column and see what that does. Use great caution when doing this, though, and be sure to wipe any residue from the outer rim of the column before replacing so that frit sits absolutely flush with the bed before putting the nut back on. If not, you can potentially introduce dead volume and have other issues like broadened peaks and tailing. Not everyone recommends doing this, but it sometimes would do the trick for me.

Though a pain, o-phosphate just seems to require frequent recalibrations due to losses. Replacing the guard column can help as well, while still sparing your analytical column.

In my experience, phosphate works a hell of a lot better on a carb/bicarb system than a system with a KOH eluent generator, so you at least have that in your favor. I gave up even trying to run for phosphate on our IC systems with hydroxide eluent at my old lab.