Chromatography Forum

Hi everyone

This is a hard one I'm facing. I have to perform a hydrolysis study with a compound that is soluble 4 µg/L, so following the OECD guideline my starting concentration must be less or equal 2 µg/L, LOQ mus be less than 0.2 µg/L. The same requirement applies to the hydrolysis products, one of which is benzoic acid. OK, so far so bad. The parent compound in question has a log POW of about 7, so a method for the parent should be straightforward (C18, high ACN, works well with LC/MS, although not to the required LOQ). But how can I also catch the benzoic acid? I just don't get it to ionize well, and furthermore I don't know how to elute it in the same run as the parent. Any idea? SPE is in my mind, I also tried in a first experiment, but recovery is rather bad. I used Waters OASIS HLB 3cc cartridges (60 mg) with 100 mL analyte solution.

Thanks a lot,
Jörg

PS.: The other theoretical hydrolysis products are somewhere in the range between benzoic acid and the parent, regarding their polarity and pKa.

Did you try microwave to increase solubility?
Maybe a HILIC column is an alternative.
Good luck.

Increasing solubility is not an option: The water solubility is determined in a separate study, and this value is what the next steps are based on.
Personally I have only bad experience with HILIC so far. And with HILIC I think I would face the same problem as with RP: One compound is not retained at all, the other strongly

Thanks anyway
Jörg

Benzoic acid has a pKa of ~4.2. I'd suppose it should be no problem to get enough retention for it on a plain vanilla C18 if you use an acidic mobile phase with a pH below that pKa (let's say a pH of 2-3).
Concerning the wide range of polarities of your analytes a gradient run seems advisable...

HPLC retention/assay of benzoic acid (such as from sodium benzoate preservative) is straightforward on RP-18 using acidic mobile phase.

HPLC retention/assay of benzoic acid (such as from sodium benzoate preservative) is straightforward on RP-18 using acidic mobile phase.

Yes, in principle it is. But I need an LOQ of about 0.2 µg/L, this is not possible with UV in a MS compatible acidic eluent system (I need MS for the parent compound). Using formic acid, I did not manage to get benzoic acid to ionize well. Any other ideas?

just some thoughts

- what column diameter are you using? using a smaller column should give higher peaks = better LOQ
-- same with gradient > isocratic

- you say with formic acid the ionization for benzoic acid (BA) was not enough. But the retention was ok? have you then tried using other ms compatible acids which gives better ionization while still protonating the BA? e.g. acetic acid (I don't have much praxis on LC/MS yet)

- just for the BA: maybe the use of a anion exchange SPE cartridge can give better recoveries for that (application with high pH). (not taking into account the other substances) maybe two runs requierd...
-- and as the elution from the ax-SPE could be done with highly aqueous solvent, you should be able to inject a larger volume without disturbing the chromatogram

- with the previuos SPE cartridge: did you acidify the application solution so to get enough retention on the HLB? Maybe a "generic" C18 cartridge would be an alternative; application at low pH

- what about the ionisation mode? does pos vs. neg ESI makes a different in sensitivity of the BA signal?

As my customer is in some sort of hurry I postponed the benzoic acid issue. In the first step, quantification of the parent is sufficient (Tier I in OECD 111 guideline). I have sufficient recovery of my parent with dichloromethane extraction, I will try out benzoic acid when the actual study is running.

I'll keep you updated and ask for more assistance later

@hollow:

- you say with formic acid the ionization for benzoic acid (BA) was not enough. But the retention was ok? have you then tried using other ms compatible acids which gives better ionization while still protonating the BA? e.g. acetic acid (I don't have much praxis on LC/MS yet)

There is my problem: Enough formic acid gives reasonable peak shape (not really good, though), but the benzoic acid is neutral then, this is what I meant with low ionisation. I thought about mixed-mode, but put no effort in this direction so far. I have Primesep A and SB here to play with, but I think both are not ideal.

Jörg