Problem with reproducible peak areas with LC/MS/MS

[Yellow]

I'm injecting a standard solution on a LC/MS/MS. I've problems to obtain reproducible peak areas. RSD's are sometimes more than 50%, but only for 2 of 15 components. Any idea how to obtain more reproducible peak areas?

Thanks in advance.

[sav]

Information?
Instrument, MS-parameters, LC-parameters, analytes, internal standards, samples etc.?

[Yellow]

The instrument is from Thermo: Accela TSQ Quantum Access

The analytes are coccidiostatics. We use two internal standards (which are added before extraction). But even when we inject only standard solution, the peak areas are not stable for some coccidiostatics.

The samples are food (meat, eggs, ...) and feed.

LC-parameters:
C18 column
gradiënt starting with ACN/H2O 5%/95%; ending with ACN 100% (both with 0.1% formic acid)
injection of 10 µl
flow: 250 µl/min

MS-parameters:
ESI, each analyt: one parent ion and 2 or 3 product ions
colision gas pressure: 1.0 mTorr
capillary temperature: 360°C
sheath gas pressure: 60
spray voltage: 3500 (positive polarity) and 3000 (negative polarity)
colision energy and tube lens value: optimized automatically by software


Is this enough information?

[sav]

Not yet

Can you post names (or CAS etc.) of “some coccidiostatics” and some data:

1. Concentration (pg/ng per peak), S/N, RSD of peak areas (n=2-4). Min. 2 points (S/N=10, S/N=100 etc.).

2. Blank EIC after injection of standard (only MRM transition for quantification). Standard EIC for this analytes.

3. Are you injected each coccidiostatic individually?

Best wishes, Alex.

[sepscientologist]

Could be something coeluting with the problem analytes and causing ion suppression. A very cool experiment to check for
this is to inject matrix without the problem analyte while teeing in an infusion low concentration analyte post-column. Then
look at the normal retention time of your problem analyte to see if the signal goes up or down from the rest of the run.

[Yellow]

The analytes having problems with reproducible peak areas are narasin and salinomycine. When we optimized the colision energy and the tube lens value we infused each analyte individually. During routine operation we measure 12 to 13 analytes in one run.

The concentration of the solution injected is about 33.3 ng/ml. Of this solution 10 µl is injected.

RSD's of the peak areas are about 7% for 5 to 10 injections. However, RSD is going up to 30 or 50% when more injections are made or when you calculate RSD on 10 injections which were from a list of 50 where you pick each 5th injection.

I am not sure what is meant with EIC, but the highest value in the chromatogram when looking only at the MRM for quantification is about 9E4 for the standard and 7E1 in blank solvent.

Do you need more information?

Since all this information is about injections with analytes dissolved in solvent, containing no matrix at all, I think there are no problems with ion suppression. Is this correct?

[MaryCarson]

Those are ionophores. Are you looking for [M+H]+ as the precursor? Try looking for a Na adduct, or maybe even NH4 adduct if the column or system has ever seen an ammonium salt buffer.
Sodium is really hard to avoid in biological sample. APCI might work better as a source. Alternately, you can try adding a trace amount of Na to the mobile phase to force everything into the Na adduct and see if you get a stable response that way. There was a publication in J AOAC Int this year (March) that uses NH4 in the mobile phase and looks at the NH4 adduct.

[sav]

EIC – extracted ion current (or extracted ion chromatogram), 700>100amu etc.

I think that the best choice in this situation: internal standard (ionophore too).

I don’t think that adding the Na+ may solve this situation, but 0.1-1mM HCOONH4 ([M+NH4]+) may help.

[Sku]

Hi,

we have a TSQ Vantage and also had in march this year problems with big differences in peak-areas (50.000 and 800.000) within one sequence for the same standard. They did not really know the problem but they changed all the boards and finally after 4 month they changed the whole device. We had a loss of intensity in the TIC with polytyrosine. Maybe you could check the TIC in your device. If it's not stable you will see the graph going down.

[Yellow]

We already use internal standards. However, these are added before the extraction. With internal standards, do you mean standards which are added in the vial?

We already performed an experiment with addition of Na+ and NH4+ (separately). The RSD's didn't get better. However Na+ and NH4+ were added in the vial and not in the mobile phase. I was always told buffers are not possible or not a good idea in combination with LC-MS. Is this not correct then? Should we take certain precautions when adding Na+ or NH4+ to the mobile phase, or are these concentrations too low to cause problems?

[sav]

1. The IS using for calibration of analytical system [(Qstd/Qis) = K*(Astd/Ais), etc.]. If you use IS (ionophore) for this pair of analyt: delta(Astd)/delta(Ais) = const (as a rough approximation).

2. The NH4+ must be in the eluent. Use volatile salts (ammonium acetate etc.).
For Na+: sodium formate or acetate etc.

[dhan]

I tend to ask 'whether your autosampler' capable of injecting exact volumes each time.

I had mine 'Accela ALS' could not perform reproducible injections. After six months or more struggling, I got 'open acela' (CTC type) and now doing fine.

Fortunately I also have PDA that has given additional advantage that to compare results with MSD response. I tested with caffeine solution.

Dhan

[kmchale]

Owing to the fact that only 2 of the 15 compounds are showing the high %RSDs, this is likely a chemistry phenomenon, not a triple quad problem.

Have you investigated the possiblity that narasin and salinomycin are either (1) are being lost during sample processing or sample transfer, or (2) are being converted to other compounds over time? For example, these coccidiostatics are similar in sturcture to tetracyclines, which bind strongly to metal ions and glass. Also, chlorotetracycline converts to 4-epi-chlorotetracycline at pH > 7 in a matter of hours.

I agree with the other authors that you should try adding ammonium acetate or ammonium formate, at a concentration of 1-10 mM, to both your sample solutions as well as the aqueous mobile phase for this type of LC/MS/MS assay. I would also investigate preparing your sample solutions at different pH values and repeat your LC/MS/MS experiments. Finally, try running a time-course study for narasin and salinomycin using a Q1 full-scan experiment to see if these compounds are converting to a structural analog or decomposition product.