Injection of Calibrant at the start of an LCMS run

[arnabroy81]

So, want to have calibrant spectra inside my LCMS data. So that I can use it for internal calibration during data analysis.

But I don't have any set up for automatically injecting calibrant at the start.

Is it okay (publishable) if I manually inject the Na-formate just at the start of my LCMS run? Anyway, first few minutes are going to be void volume. Then I can connect the LC port to the ESI.

Any suggestion will be highly appreciated.

[HPLC chemist]

No! If you are in the pharmaceutical arena your method MUST comply with USP <621> (or equivalent) which is science based. This is also a 'good' business strategy!

Most companies and their equipment automatically inject the standard (your calibrant) into the MS detector along with the standard.

[lmh]

I'm puzzled. HPLC chemist, I'm not sure if you're referring to an internal standard used in calibration of a quantitative method? I'm guessing that the original poster is referring to a sodium formate calibration solution being used to provide mass calibration of an accurate mass instrument?

It is certainly possible to publish accurate mass MS data where calibration is by inclusion of a short infusion of sodium formate at some convenient point in the run, and this infusion is used to calibrate the remainder of the run. I think Bruker used to do it that way (I have no idea what they currently recommend). The question is how fast your MS drifts. If it drifts very fast, then calibrating using an infusion from some minutes previously wouldn't be as reliable as using continuous alternation of lock-mass and sample, in the way that Waters would. The important thing for publication is whether your mass accuracies remain within acceptable bounds - usually 5ppm.

The remaining question is whether it's OK to infuse sodium formate manually before quickly reattaching the tubing, rather than doing this automatically. Since you're not altering the pre-column plumbing in any way, the retention times cannot be changed by your activities. In an academic environment I think it would be grossly unfair of a referee to refuse your paper for doing this manual infusion, but you should be honest about it.

It might be easier to automate this system than you think. Do you have a divert valve on the MS? If so, you can plumb it to do loop injection, with a fairly big loop. For most of the run, you keep the loop out of the flow-path, and during this time you can fill the loop with sodium formate, either manually or by a very large syringe pump working at a very low flow. At a convenient blank bit of baseline, either before your retained peaks, or during re-equilibration of the column, switch to the "inject" position, so that the flow takes the loop's contents to the MS. The switching is, of course, automatic.

[HPLC chemist]

Perhaps I should not be so hasty in my reply!

A reference standard is used in a method to demonstrate the instrument's ability to identify and quantitate a particular target molecule. This must comply with USP <621>.

A calibrant can be used on a LC/MS/MS detector for peak identification and m/z ratio calculation (this is the equivalent of wavelength accuracy in a PDA/DAD). In this case the MS can be detached from the HPLC. This would be an 'Operational Qualification (OP)'.

Finally, the detector is reassembled to the HPLC and a target molecule of known concentration injected, identified, and quantitated per that companies SOP. This could be considered a PQ or 'Performance Qualification'.

My preference (obviously) is all 3.

[lmh]

No worries! I still don't know whether I'm interpreting the question correctly. It's unfortunate that we use the word "calibration" to cover both calibration with standards of the analyte, for quantitative work, and to cover mass calibration, which is a completely different thing. And I get quite frightened about whether what I do is reputable or lax.

Incidentally, to the original poster: if your instrument doesn't drift too badly, but you'd like to correct for very small drifts and get the best possible masses, does it have the capacity to use a single mass in the signal as a lock-mass, or to polish up your calibration post-run? Thermo call this a lock-mass and provide the functionality on orbitraps (haven't had a chance to use one for a while, but it was a tiny button marked with an "L"). Waters provide it on the Q-ToFs, but also have a highly technical way to add the lock mass deliberately at intervals through the run. Shimadzu provide a fairly simplistic version on their IT-ToF, where it's called "mass compensation". If you have this functionality, then look to see if you can find a friendly phthalate (391.2843; 413.2662) that's in your solvent already, and use it as a lock-mass/calibration-polisher).

[arnabroy81]

HPLC Chemist and Imh, you guys are awesome. I get to learn a lot of things from the discussion.

Although here I was trying to use a calibrant (Na-formate), the options you gave seems worth trying. Especially the lock mass one. Yes, I see a way in the DataAnalysis software that allows me to incorporate a lock mass calibration during data analysis.

I didn't find any of the pthalate peaks in my spectra. Is there a good way I can inject a lock mass after the LC-column? May be connect a T-joint and inject the lock mass? Which molecules will be good for a lock mass calibration? Reserpine? Or other standards?

Till now I was just calibrating the data with the calibrant injected at the start. Now I can experiment with lock mass.

[Rndirk]

What kind of MS are you using?

[lmh]

I'm not sure what I'd add... it probably depends on what you're measuring, and it hurts, adding a potential contaminant deliberately to a buffer! In Orbitrap world it doesn't seem to matter too much if the lock-mass isn't very close to what you're measuring. In Shimadzu instruments (but not many people have them) it probably does. I'm not sure about other ToFs, such as Waters.

[arnabroy81]

What kind of MS are you using?

I am using a Q-TOF from Bruker

[jenche]

I've been a Bruker QTOF user for 5 years now (both on maXis 4G and now maXis II, but what I'll say is also relevant to the impact and the compact instrument).

I've never had to use a lock-mass on these mass spectrometers as they don't drift that quick (especially on the maXis/impact II which are really stable) (assuming the room temperature doesn't change too much during your analysis...).

Best way to calibrate the mass spectrometer per analysis is to fill a syringe with your calibration solution and plug it to the divert valve. The idea here is to fill the loop with the syringe, and to switch the divert valve at the beginning of the analysis. You can do that by adding a segment in your otofControl method and to use the divert valve option (switch it to Waste or Source, depending on the way you've plumbed it). In recent version of otofControl, you can define a segment as "calibration segment" with can then be used by DataAnalysis to recalibrate your data using the "Automatic internal calibration" function.

If you want to do that automatically for many analyses, you can use a syringe driver that will fill the loop during the analysis.

I can't really explain more because I'm not currently near my instrument and I don't remember every details, but from what I remember this kind of use of the divert valve if explained in the instrument manual. If you need more help, let me know and I'll have a look.

Also, you might want to use an additional lock mass if you have very long run-time of if the temperature condition of the room make the instrument drift, but be aware that it can decrease your signal.

[arnabroy81]

Thank you very much Jenche. I'm currently doing exactly what you have mentioned. I'm keeping a 0.5 min calibration segment at the start of the run and using that for calibration in DataAnalysis.

The mass drift you have mentioned, however, is a concern for me. I'm not sure what is the cause but, i'm noticing a mass drift of about 0.02-0.1 Da shift in mass measurement within few seconds. And if I keep the instrument in standby for two days (weekends), It gets really hard for me to calibrate the machine. It drifts too much!

I'm quite sure this is not normal. Can someone support me here?