Analytical Challenges in Quantifier Ion Selection Due to Mass Error Variability

Hi, I am currently developing an LC-HRMS method for the analysis of an NDSRI in a powder-for-injection formulation. The compound belongs to the penem class.

For sample preparation, I am using water as the diluent. During mass spectrum interpretation, particularly while selecting the quantifier ion, I observed that the parent ion in the sample shows a relatively higher mass error (~6–7 ppm), though it is still within the acceptance limit (±10 ppm). In contrast, the NDSRI reference standard exhibits a much lower mass error (~1 ppm).

Based on prior experience, I initially suspected a matrix effect. However, the formulation contains only a single excipient—sodium carbonate—which makes the situation less straightforward.

I would appreciate your insights or suggestions on possible causes for this observation. @Phil @Naiffer_Host

Have to tried to measure NDSRI reference standard diluted in sample matrix (placebo) ? Do you get the same shift here ?

3 Likes

Hi @Phil thanks for your reply. yes i did. i prepared the NDSRI in the placebo(i.e sodium carbobnate) . but the NDSRI peak was not obtained due to degradation at higher pH.. But I spiked the NDSRI in the sample at 100% level and i found the same shift there too.

2 Likes

Okay, if the same shift is observed, it is most likely a matrix effect.

Sodium carbonate may sound simple, but depending on the concentration, it introduces a significant ionic load into the system.

In addition, the adduct pattern may differ. Many nitrosamines tend to form [M+Na]⁺ adducts. If this is the case, you could consider adding a small amount of sodium (e.g., Na⁺ in the eluent) to stabilize the your system and reduce the discrepancy between samples and standards.

2 Likes

Thanks @Phil.. in my case no sodium adducts are formed. The mass of the parent ion differs in the last two digits after the decimal point. for example, if my exact mass is xxx.1589. my observed mass is xxx.1579 or xxx.1525 like that. So, we are going with the daughter ion as the quantifier ion. Another important thing is that the mass spectrum of the parent ion is showing three peaks with 0.5dalton spacing. will be it the pattern of dimer?? but the mass pattern of the standard and sample are same except the parent ion mass.

@mayank.bhanti @lucas10mauriz Can you provide your expert advise here? Thanks

1 Like

The presence of three peaks with ~0.5 Da spacing indicates a doubly charged ion ([M+2H]²⁺) rather than a dimer. Since penems contain sulfur, the isotope envelope is influenced not only by ¹³C but also significantly by ³⁴S (~4.2%), which enhances the M+2 peak intensity. As a result, the parent ion does not appear as a single dominant peak but rather as a cluster of closely spaced isotopic peaks of comparable intensity. The fact that the isotope pattern is consistent between standard and sample confirms that the same ion species is being observed. However, the shift in measured mass (last decimal digits) in the sample is most likely due to: Centroiding differences within the isotope cluster, Matrix effects distorting the isotope envelope, Possible minor co-eluting interferences affecting peak shape. Given this behavior, relying on the parent ion for quantitation becomes less robust. Therefore, selecting a daughter ion (fragment ion) as the quantifier is a better, particularly if it shows better mass accuracy, lower variability across injections, and reduced matrix interference.

3 Likes

I’d like to add a few additional considerations that may be helpful for the next stages of method development.

1) On the mechanism behind the mass shift:
I agree that the 0.5 Da pattern clearly supports the presence of the doubly charged ion, [M+2H]²⁺. One aspect that may complement your interpretation is understanding why this behavior differs between the standard and the sample.

The literature on solvent effects in ESI (Effect of eluent on the ionization process in liquid chromatography–mass spectrometry) shows that more polar and alkaline media stabilize multiply charged species in solution more effectively, shifting the charge-state distribution toward higher charge states prior to gas-phase ion emission. While acidic conditions generally favor [M+H]⁺ formation in positive ESI mode, the alkaline environment created by sodium carbonate in pure water may instead favor [M+2H]²⁺, particularly for a compound with multiple ionizable sites such as a penem. This would explain why the standard, if prepared in a neutral or slightly acidic diluent, behaves differently from the sample, a mechanistically expected outcome rather than a random artifact.

The same work also reports that electrolyte concentrations above 10 mM can already produce measurable ion suppression in ESI. The sodium carbonate present in the formulation may well exceed this threshold in the sample solution, introducing an additional ion suppression effect on top of the charge-state issue itself.

2) A point to consider regarding the daughter ion as the quantifier:
Before establishing the product ion as the quantifier, it is worth confirming an assumption that is not yet entirely clear: are the standard and the sample actually generating different charge states, or do they both exhibit the same 0.5 Da isotopic spacing?

If the standard also ionizes predominantly as [M+2H]²⁺, then the likelihood of differential fragmentation under CID is minimal, if not nonexistent. On the other hand, if the standard is observed as [M+H]⁺ with a 1.0 Da isotopic spacing, then this point certainly deserves verification: precursor ions with different charge states experience different effective kinetic energies at the same nominal collision energy, which can lead to distinct fragmentation profiles and potentially compromise the consistency of the transition pair between standard and sample.

Therefore, the first practical step is simply to inspect the spectrum of the standard using the same level of zoom that was applied to the sample cluster and confirm the isotopic spacing. The answer to that question will determine whether this risk is real or not.

3) A practical path forward:
One approach that could simultaneously address both the charge-state distribution and ion suppression would be to prepare the samples in a diluent with controlled pH.

For example, 10 mM ammonium formate adjusted to approximately pH 3.5–4.0 instead of pure water. This would likely promote a more uniform ionization behavior between standard and sample, favor the formation of [M+H]⁺, and reduce the impact of carbonate species at the ESI droplet surface.

Following this adjustment, a formal matrix factor assessment would also be advisable to confirm, quantitatively, that the observed effects have indeed been mitigated.

I hope these thoughts are useful as the method development progresses. I’d be happy to discuss any of these points further if that would be helpful.

1 Like