Drift and change of response factor (aminosalicylic acid)

[Mattias]

I would really appreciate some help on this, it is driving me crazy....

I am analysing 5-aminosalicylic acid (5-ASA) on a mixed-mode column using TFA/water/acetonitrile as mobile phase. The sample is dissolved in an EDTA solution in water (pH about 7). The reason for using EDTA is that 5-ASA is very easily oxidised, and EDTA makes the solutions stable for weeks.

The problem is that when I start a new sequence, I get a quite a lot of area drift (increasing areas). It can take up to 10-15 standard injections before the areas become stable. Another strange thing is that the response factor of the standard is quite different in different sequences (using the same instrument and mobile phases). It can shift plus/minus 5% from one day to another.

I have performed quite a lot of trouble shooting already:
- I replaced the column with some PEEK tubing = same area drift observed
- I have used a metal-free LC-system = same area drift observed
- I have tested to siliconise the needle and the LC-vials. It seemed to increase the response factor, but it didn't remove the drift.
- The area drift seems to get worse if I inject the standard from the same vial, compared to performing one injection/vial. The drift is however too large and fast to be explained by evaporation.

All ideas are welcome!!

[zokitano]

Hi Mattias,

What concentration of EDTA are you using in the final injection sample solution? Did you try to optimize the EDTA concentration in your samples, as you noticed (and from my own experience) that 5-ASA is a good complexating agent (ligand) especially toward the residual heavy metals found in the columns or LC solvents?
From the text I can conclude that there are probably some active sites (heavy metals embedded in silica support probably) that need to be "saturated" or "bound" with your analyte before you get stable peak areas.

Another possibility to investigate is to put some residual concentration of EDTA (micromolar) in your mobile phase (aqueous part) and to see if this addition will improve the consistency of your analyte peak area. If you're using UV detector probably it would not be a problem to use EDTA mobile phase, but if you use MS detector, there might be some additional suppression effects from the EDTA mobile phase (that you'll have to investigate) or you have to clean the ion source more frequently to keep the consistency detector response for your analyte. But I don't think that EDTA will harm the MS ion source.

Hope this helps.
Regards

[Mattias]

Thank you very much for your reply!

The sample solvent is a 5 mM solution of tetrasodium-EDTA salt. If I dissolve 5-ASA to 1 mg/ml in this solution, I get a pH of about 6.5 - 7.0.

I am only using UV-detector, so I can defintely try to add EDTA also to the mobile phase. I will try that today, and post the results!

[Vlad Orlovsky]

saturating/washing your system with EDTA might address your issues. EDTA can from strong complex with metal parts and then behave like an ion trap for your compound. I would also try other anti-oxidizing agents (ascorbic acid?) instead of EDTA. You should not have retention of these compounds on your mixed-mode column because they are hydrophilic. I also would look at things like needle wash, source of vials and other things related to your system.

[Mattias]

Last night I added 0.5 mM of EDTA to my aqueous part of the mobile phase. The system has been running all night, and I can see the same area drift as before...

I think the problem is that EDTA is not active at the pH of my mobile phase (pH about 2). The EDTA is there but it does no good!

Are there any metal chelators that are working at a low pH?

[zokitano]

Last night I added 0.5 mM of EDTA to my aqueous part of the mobile phase. The system has been running all night, and I can see the same area drift as before...

I think the problem is that EDTA is not active at the pH of my mobile phase (pH about 2). The EDTA is there but it does no good!

Are there any metal chelators that are working at a low pH?

I used EDTA in the injected sample at lower (micromolar) concentration together with RP mobile phase pH=3, and it worked well for my problematic analytes. I had peak shape problems and senisitivy issues for some salicylic acid derivatives and some catechol group containing analytes. The addition of EDTA only in the sample solved my problems, but here in your case maybe the 5-ASA complexation is not the main reason for the peak area inconsistency.

To second Vlad's opinion, are you sure that you are not facing with a consistent carry-over of your analyte in the later injections?

[Mattias]

Without any kind of rationale, I added some EtOH to my standard and let the system inject over night.

This sequence did not show any drift at all. It appears as Vlad was right, pointing the problem towards adsorption.

However, I will need to repeat this on another system to be sure that everything is OK now.

[zokitano]

I'm glad that you made it and thanks for reporting back the results.

Cheers