LC/MS/MS (ESI)-When is it appropriate to quantify Na adduct?

[MMJ88]

I am doing LC/MS/MS method development on a variety of cytotoxic drugs. One of them, paclitaxel (http://en.wikipedia.org/wiki/Paclitaxel) tends to form a sodium adduct. I ran a syringe pump straight into the MS to investigate, and it occurs with ACN/water and MeOH/water (all LC/MS/MS grade reagents). I also tested adding formic acid at 0.1% and 1.0% (thinking increasing [H+] could shift the balance away from the sodium adduct. None of this made much of a difference.

In actually doing some runs with chromatography and looking at the data, the sodium adduct has a S/N ratio ~20 greater than the hydrogen adduct. Is it appropriate to quantify based on the sodium adduct?

Another factor is that I have other analytes that form primarily the molecular ion, so I'm hesitant to add sodium acetate to "force" the equilibrium of paclitaxel.

Any insight would be much appreciated. I've been away from LC/MS for a few years and am just getting my feet under me.

(Instrument is an Agilent 6410 with ESI source).

[bunnahabhain]

There are some molecules that are not forming the [M+H] ion at all, I have made this experience, too. Instead of adding sodium acetate, you might want to try out ammonium formate to see if you can get the ammonium adduct. You could then add traces of ammonium formate to your eluent system (if you don't need it buffered, 1 mM is sufficient) because it won't clog your ion source, while sodium salts will do. Relying on the trace amount of sodium that is almost always present, is not advisable, it might change from batch of water or sample. Having a defined amount of ammoinum present in the eluent might make quantification a bit more reliable.

Just my thoughts about this issue, YMMV

Jörg

[carras]

In my experience another drawback of sodium adducts is that they don't fragment in the collision cell of triplequads so that MS/MS determinations cannot be performed.

[MMJ88]

Yes, the fact that sodium content in mobile phase is a contaminant and thus not safely assumed to be constant is the main problem I see.

I will take you suggestion about ammonium formate into account.

There are some molecules that are not forming the [M+H] ion at all, I have made this experience, too. Instead of adding sodium acetate, you might want to try out ammonium formate to see if you can get the ammonium adduct. You could then add traces of ammonium formate to your eluent system (if you don't need it buffered, 1 mM is sufficient) because it won't clog your ion source, while sodium salts will do. Relying on the trace amount of sodium that is almost always present, is not advisable, it might change from batch of water or sample. Having a defined amount of ammoinum present in the eluent might make quantification a bit more reliable.

Just my thoughts about this issue, YMMV

Jörg

[MMJ88]

yeah, I've seen this in other cases, but in this case it does make a nice MRM transition.

In my experience another drawback of sodium adducts is that they don't fragment in the collision cell of triplequads so that MS/MS determinations cannot be performed.

[MMJ88]

You could then add traces of ammonium formate to your eluent system (if you don't need it buffered, 1 mM is sufficient) because it won't clog your ion source, while sodium salts will do. Relying on the trace amount of sodium that is almost always present, is not advisable, it might change from batch of water or sample. Having a defined amount of ammoinum present in the eluent might make quantification a bit more reliable.
Jörg

Hi Jörg,

I tried this. Didn't have any ammonium formate on hand (ordered some though), so I added a bit of ammonium hydroxide to 0.1% formic acid. Should function similarly. This solution appeared to help quite a bit for M+H formation vs. M+Na. I didn't see the ammonium adduct though, it just seemed to shift the adduct formation away from Na. Is this an expected result?

[bunnahabhain]

You could then add traces of ammonium formate to your eluent system (if you don't need it buffered, 1 mM is sufficient) because it won't clog your ion source, while sodium salts will do. Relying on the trace amount of sodium that is almost always present, is not advisable, it might change from batch of water or sample. Having a defined amount of ammoinum present in the eluent might make quantification a bit more reliable.
Jörg

Hi Jörg,

I tried this. Didn't have any ammonium formate on hand (ordered some though), so I added a bit of ammonium hydroxide to 0.1% formic acid. Should function similarly. This solution appeared to help quite a bit for M+H formation vs. M+Na. I didn't see the ammonium adduct though, it just seemed to shift the adduct formation away from Na. Is this an expected result?

You never know what you can expect from addition of ammonium, but it's always worth a try. Start playing around a lot with needle voltage and cone voltage using flow injection and see if you can push your M+H further up.
Good luck

Jörg

[Camisotro]

Hi Jörg,

I tried this. Didn't have any ammonium formate on hand (ordered some though), so I added a bit of ammonium hydroxide to 0.1% formic acid. Should function similarly. This solution appeared to help quite a bit for M+H formation vs. M+Na. I didn't see the ammonium adduct though, it just seemed to shift the adduct formation away from Na. Is this an expected result?

That's quite typical, yes. Remember ESI is dependent on low levels of ions already in solution to kickstart the ionization process. If you have an analyte that likes to stick to cations, and trace Na+ is the only cat in town, you will get Na+ adducts which don't like to break up in the gas phase. If Na+ is outnumbered in solution by intentional volatile ions such as ammonium and formate, the ionization process no longer hinges on the Na+ and you'll see more of the protonated molecule. Ammonium falls off adducts in the gas phase much easier than sodium.

What's important is you've shifted the equilibrium away from an ion dependent on factors beyond your control, to an ion dependent mostly on controlled factors.

[MMJ88]

Hi Camisotro,

Thanks for the insight. That is basically what I have found in my experimentation.

One thing still remaining to figure out is I only get this benefit when directly infusing from a syringe into the MS. When the same sample is injected through the LC side, I see 99% sodium adduct. So I guess I'm looking for sodium sources in the LC side? It's quite frustrating. I'm using all LC/MS/MS grade reagents. I washed my solvent bottles thoroughly. Anyone else have any ideas?

[bunnahabhain]

There is no way to avoid sodium. The specification stated in Chromasolv LC/MS and LC/MS ultra grade is ≤0.1 mg/kg sodium. Additionally, as soon as your mobile phase sees glass vessels, some sodium traces are extracted and available to your analyte. This is obviously enough to saturate your analyte. The only thing you can try (but did not work out for you) is, to try to shift from an adduct you can't control the concentration (Na+) to one you can control (NH4+).

[MMJ88]

Hello again, Jörg.

Right you are - the small amount of sodium extracted from mobile phase reservoirs or sample vials does seem to be enough to saturate my sample. Unfortunately the compound in question (paclitaxel) seems to have quite an affinity for the sodium adduct. Looking at one other compound (cyclophosphamide) I can see the 1mM ammonium formate has an effect just as you (and others) predicted - the cyclophosphamide shifts strongly away from the sodium adduct and the H+ adduct is 3 or 4-fold larger. But paclitaxel is hardly affected.

I read in Regulated Bioanalytical Laboratories by Michael Zhou that carboxyl ester groups may be responsible for binding the metal ions - this would make sense, as paclitaxel has an abundance of such groups.

I have also tried 5mM ammonium formate, and just for kicks, tried 1mM and 5mM also in the sample solvent. None of that really made a difference. I'm getting to the point where I may be wasting too much time on this, but any further ideas are still welcome.

Thanks a lot for your replies.

[emeulman]

Most of the people analyze paclitaxel with The MRM transition of m/z 876.38-308.00 for paclitaxel. So this is a sodium adduct. Apparently one easy to fragment. Just use 0.1% formic acid in water and 0.1% formic acid in acetonitril as mobile phases.

[meff.c]

Hi MMJ88

I have exactly the same equipment and problem with Sodium adducts. Did you fix the problem? I would appreciate your help.
bye.
Meff

[MMJ88]

Hi Meff,

As emeulman alluded to earlier, I found a lot of papers using the 876 parent ion for paclitaxel. Regardless of what I tried (mainly different mobile phases), the Na+ adduct was by far dominant over the H+ (using ammonium formate did make a difference in the adduct balance, but only for the other compounds, paclitaxel was basically unaffected.). The calibration and QC samples were fine and I saw the same ratio in spiked vs. environmental samples. So I just went with the sodium adduct for quantitation.

If you're interested, you can read the paper here: http://link.springer.com/article/10.100 ... 015-8868-y.

Cheers.

[Klaus I.]

Beside of the also possible pronounced affinity to sodium there is an alternative explanation for the missing [M+H]. Maybe all of the formed [M+H] are fragmented in source. Sodium-adducts are often more stable which may also effects the intended fragmentation for MRM or MS/MS negative.
Reducing temperatures and voltages in the source and cleaning of the source might help in this case.
Nevertheless I have no personal experience with Paclitaxel, but it looks quite fragile on the first view.