Strange retention mechanism

[John G]

I developed a LC-MS esi neg SIR method for determination of chloroacetic acid. The method uses an Xterra 2.1x150 mm MS C18, 5 µm column and the mobile phase is 95% 2mM acetic acid : 5% ACN at 250 µL/min. The chloroacetate eludes at ~9 min. I was tweaking the method and increased the ACN content to 10%, 15%, 20% and 25%. Each increase in ACN content increased the retention of the ion, ~22 min. at 25% ACN. If effect was caused by change in pH due to the increase in ACN then that would cause a decrease in the retention of a acid analyte in theory. Does anyone seen something similar or have an explaination for what I am observing?
John

[ConstantinS]

John, maybe it is not Cl-acetic acid? k' is about 4-5 on 5% AcN eluent - its too much!
Is this peak broad or asymmetric?

[John G]

The chromatographic peak is 0.6 min. wide at 5% of height and completely symmetric. The resolution of the the m/z 93 ion is set relatively high with a ion peak width of 0.5 Da. I have run several sets of standards covering 4 orders of magnitude and shown excellent linearity, r squared = 0.9998. Spiked sample solutions for different samples all showed good results, so I don't have any reason to suspect that I am not looking at the correct peak. Other then the strange retention behaviour.
John

[MG]

I analyzed this compound and other haloacetic acids by LC/MS/MS, using a Luna C18(2) with an acidified water/methanol gradient. I observed reverse-phase behavior. So all I can say is, I agree that your results sound strange!

[Bill Tindall]

I don't know about this specific column but I have seen anion exchange behavior in supposedly reversed phase columns. Under these conditons the chloroaceitc acid will most likely be mostly dissociated into an anion. What you are seeing is consistent with this sort of retention. Add a pinch of formic to the higher ACN run and see if retention time decreases. Phosphoric would be better for this experiment but then you might have to detect by UV.

[HW Mueller]

Anion exchange was my first thought also as I have seen this at low pH. It could be that there were SiOH2+, but the literature seems ambivalent on this.

It would be surprising that the decrease of acetate due to the increase of ACN would be enough to lower its competition for stationary + to that extend? The thought that the ACN causes higher acidity even though acetate decreases also seems unlikely? Thus the suspicion is strong that there is an entirely different cause behind this, some sort of undetected mistake?

[ConstantinS]

Maybe its not XTerra?
Something like mentioned not long ago OmniPac or Primesep?

[Victor]

Why not try this analysis on another column (especially a column from a completely different manufacturer-like the Luna column suggested above) and let us know what happens?

[Victor]

Also, the other correspondents seem to have assumed that there is no acetic acid in your acetonitrile bottle i.e. that the overall concentration of acetic acid drops as you increase the ACN content. Is that right? Maybe you should try maintaining the concentration of acetic acid constant and see what happens.

[John G]

Thanks for your everyone's input so far on this issue. I have another column to try and will try methanol instead of Acetonitrile to see what effect that has as well. I don't know how much effect keeping the acetic acid concentration constant will have since it is so low, but if I have time I will try that too.

I was wondering about the MS/MS method that MG developed. Chloroacetic acid doesn't fragment very well on our instrument and the only fragment it gave was the chlorine at 35. I wanted to do MRM analysis but had significantly less sensistivity and ended up using SIR to meet the required LOQ. MG what transistion did you monitor in your MS/MS method for the and if possible what levels (LOQ) could you see for chloroacetic acid.

Thanks to all
John

[MG]

The daughter I used was 49, but chloroacetic was used as an internal standard for the other acids, so I did not evaluate an LOQ/LOD.