while working with the ESI/MS (LCQ Advantage)i have been trying to do quantitative work on substituted diphosphonic acids. Their structure contains two ionizable phosphonic acid groups and C8 alkyl chains attached to the phosphorus via oxygen. The compounds are very hydrophobic. Using a C18 column and a 70:30 MeOH/20mM ammonium acetate mobile phase I am unable to get rid of the extensive peak tail.
Is this due to the long alkyl chains?
Does the fact that diphosphonic acid is a strong metal ion chelator complicates the chromatography?

I greatly appreciate any suggestions.

Hello Peter,

I suppose that your analysis is performed in negative ionization.
I have used a mobile phase containing 1-10 mM formic acid and acetonitrile for the separation of sulfonated metabolites of acetanilides herbicides and for other acidic (carboxilic) herbicides. The separation work well with a Purospher Star C18 endcapped column and negative ionization ESI-MS.
The slightly acidic condition provided by formic acid at low concentration is probably sufficient to suppress the chelating property of your analytes
I hope this note is useful for you.

Thank you for the response. I looked at some properties of the purospher silica. Is it the low metal content that prevents the chelation and the peak tailings? If that is the case I will probably get one. Thank you,
Peter Zalupski

The cause of your severe tailing problem is most likely chelation of your analytes with trace metal components. It is helpful to utilize columns based on high purity silica in this sort of application but this alone is not sufficient. Columns are readily recontaminated with metals from the eluent, the sample, the solvent filter in the eluent bottle (assuming it's made of metal) and the column hardware (especially the frit). You will have to take preventive measures to stop all these sources of metal contamination from causing problems with your assay. Your best course of action, in my experience is to operate at a pH low enough to prevent your analyte from being an effective chelator. In the case of phosphonates this will require the pH to be less than 3 and probably better still less than 2. Otherwise, if this is not feasible another approach is to spike the sample with a more potent chelator, preferably with a retention time shorter than that of your analyte.

Hi, Chris. Thank you for your suggestions. Low pH is probably a way to go. I will loose sensitivity though, using the ESI(-). Should I switch to the positive ionization mode?
Thank you,
Peter Zalupski

how about adding a low concentration of a chelator like EDTA to the mobile phase?

To Chris Pohl:

Can we remove metal ions from eluent and the solvent filter in the eluent bottle (assuming it's made of metal) by using an ion exchange guard column? If yes, what would you recommend? We can remove metal ions in the sample by adding some strong chelating agent in the diluent. There are columns available that use plastic inner sides and frits.

Hi Peter
you are right about characteristics of purospher silica.
It is a good practice to try both ionization mode (positive and negative) for almost any analytes when work with ESI interface.
Prediction on behavior for analytes ionization based on their structures will fail in certain cases with regard to ESI interface. After evaluation of response and pattern of fragment eventually generated you can make a choice. Look for an example of pH optimisation for acidic analytes, at Journal of chromatography A, vol. 878 (1)(2000) pp. 87-98
sorry for my english
regards
Anonymous #1

I'd like to thank all of you for the interest and valuable suggestions. I plan to inject an EDTA sample to observe whether similar tailing occurs.
With my best regards to all of you,
Peter Zalupski

To Anon (Wednesday 3:43 a.m.)

If you place a trap column ahead of the injection valve you can prevent metals from entering the column except for those which are present in the sample. Generally, it's best to use a chelating trap column rather than a simple cation exchange trap column in order to prevent any metal ions from leaking from the trap column as a consequence of the ions present in the mobile phase. The best choice is a trap column based on iminodiacetate. Dionex sells one for this purpose: MFC-1 #037017 (more information at: http://www.dionex.com/app/tree.taf?asset_id=11801). However, another thing to keep in mind is the need to remove any metals which have accumulated on the column prior to the aforementioned preventative measures. Typically when trying to assay chelating agents under conditions were the chelating agents are active, it is beneficial to make replicate injections of EDTA until the EDTA peak looks "normal". Of course as suggested above, adding a chelating agent to the eluent can work very effectively at masking the problem. But just make sure the chelating agent you choose has minimal absorbance at your detection wavelength and that the metal complex formation constants of the chelating eluent additive are greater than that of your analyte. Otherwise, your analyte will simply "steal" metals from the chelating agent in your eluent. An analogous solution to the problem is to add the chelating agent to your sample, which eliminates the requirement that the chelating agent (or the resulting metal complexes) not absorb at the detection wavelength, provided they don't coelute with your analyte. Of course, in this latter case it is still essential that you make sure the stability constants for your chelating sample additive are greater than those of your analyte. There are a few sources of this information. The Lange's handbook has quite a few. The following link can get you to several sources of information: http://www.stanford.edu/~cpatton/constants.htm
The most complete compilation I'm aware of is: Martell and Smith.