LC/MS/MS Quadrapole Sensitivity Issues

[BHolmes]

Hello All,

Our lab currently uses a Waters Acquity UPLC coupled to a Waters Xevo TQ-S Triple Quadrapole MS detector for analysis of pesticide residue in food. Sample prep is QuEChERS. This setup has been in use for about a year now and we are starting to notice a loss in sensitivity for our samples. Waters has completed a thorough cleaning of the entire system, and by their tests its okay. But with our samples, its acceptable but not the best that we have seen.

To give examples, Diazinon is our most sensitive compound normally at around 2,000,000 response for a 10x LOQ standard. But lately is has been only half that, at around 1,000,000. Isoprocarb is our process control and is normally around 170,000 but lately is has only been around 50,000. The integrity (repeatabiltiy) of calibration curves (we bracket samples with standards and use all standard injections for calibration curve) is okay and the R value passes. Its just the loss in response.

Any suggestions on how we can troubleshoot this further or has anyone with similar instrumentation and application experience the same sensitivity problem?

Thank you!
Brittany

[Gaetan Glauser]

Hi Brittany,

I don't really know about the tolerance that Waters has set for sensitivity tests, but if the tests pass, your MS system should do not too bad. What about the noise, is it also twice lower?

I guess they used direct infusion to test sensitivity. Can you ensure that the sensitivity issue doesn't come from the UPLC system (problem with injection, column wearing that may lead to wider and therefore smaller peaks - what about peak areas?)? Or the mobile phases? Or new calibration solutions that may be slightly different?

We've had several times sensitivity issues on our systems. Sometimes we could not entirely get back to usual values after thourough cleaning, but after several rounds of cleaning and sample runs, we always obtained normal or almost normal signals (80-90%).

Finally I agree it's pretty annoying to observe a loss in sensitivity, but as long as the system is good enough for your application, you may consider it's not detrimental.

[Aviator]

Hi Britanny,

I use an Agilent system so I couldn’t give you any instrument specific advice, but what I would ask is when was the last time you tuned your instrument and perhaps most importantly, how old is the multiplier?
Cheers,

Aviator

[Gaetan Glauser]

Aviator, that's a good point but on the Waters TQ-S, I think there should be this Intellistart thing that requests you to perform a detector (electron multiplier) check every 3 months or so. The engineer that cleaned the machine certainly performed the electron multiplier check, otherwise that was a bad engineer...

[lmh]

The Waters TQ-S doesn't have an electron multiplier.

[BHolmes]

Thank you all for your quick and informative responses, I knew there was a reason I preferred this forum over others!

The engineer did perform some sensitivity tests, but we did not have the correct solvents for her to completely sign off on the result. Waters SOP called for Aceonitrile Optima LCMS grade but all we had was Optima LC grade, so she used MeOH Optima LCMS grade instead. With that solvent, the sensitivity result was not where she wanted it to be, but it was acceptable according to her. I have some of the correct grade on order and will run test to see if I can get better results. The test encompasses both the UPLC (analyzing the a series of dilutions with the UPLC) and MS (direct infusion using the Intellistart). SO it should let me know if my problem is with UPLC, if there is one - my supervisor thinks there is Once they see the best, they want it ALL the time!!


Thanks again, and I will post updates sometime next week after I have run tests with correct solvent.

[Gaetan Glauser]

Correct, sorry I confused with the SynaptG2 QTOF which holds an electron multiplier. The TQ-S has a photomultiplier.

[Alp]

Apart from your solvent issues, which may hold an answer,
I agree you should look into the multiplier. As they age you usually have to adjust a voltage parameter or such. You also may want to verify the masses your instrument sees have not shifted a little bit to what it was seeing in the past.
Another thing you may want to check is to verify that your Q1 and Q3 resolution parameters are the same. I have experience on API systems on which you can widen the resolution of the instrument for a Q1 and/or Q3 scan allowing for a slight gain in sensitivity. I am not certain your system allows this but I see no reason why it should not.

Good luck.

[lmh]

Yes the Xevo allows you to choose different resolutions (independently on both quadrupoles). You will lose some intensity in going down to 0.5Da, so if you are on the high resolution setting, that may be unhelpful. But the main problem is if you're on high resolution and the mass calibration drifts at all, in which case the signal may begin to drift significantly outside the window. Personally I haven't had the problem, but I remember it being mentioned during training. I see little point in the high-resolution setting anyway, because near-isobaric compounds need much much much better than 0.5Da if they're to be distinguished. I can't think of any two conventional compounds that could be distinguished at 0.5Da and not at 0.7Da!

[carras]

Hello Brittany,
Once we experienced a sudden loss of sensitivity, to put it mildy, with a Waters QuattroMicro. We went from seeing a chloranphenicol peak with S/N better than 30 for a concentration of 5 ng/ml (MRM) to seeing just noise. Other analytes were also affected but I focused in chloranphenicol because Waters uses it for their sensitivity check. Waters service checked the QMIcro infusing directly their standards and everything was OK, only that we could see no peak in LC/MS. Then they tried the HPLC (Alliance) but everything worked as expected. Intriguingly, if you changed the mobile phase fron 30/70 acetonitrile/water to 100% acetonitrile you got a nice peak; pity about not having any retention in the column. After quite a long time and much coming and going from Waters technical support they told us that the first quadrupole was dirty. They changed it and the signal was back to normal. Apparently they tried ultrasonic cleaning of that quadrupole to no avail.
Hope it helps

[BHolmes]

Hello Brittany,
Once we experienced a sudden loss of sensitivity, to put it mildy, with a Waters QuattroMicro. We went from seeing a chloranphenicol peak with S/N better than 30 for a concentration of 5 ng/ml (MRM) to seeing just noise. Other analytes were also affected but I focused in chloranphenicol because Waters uses it for their sensitivity check. Waters service checked the QMIcro infusing directly their standards and everything was OK, only that we could see no peak in LC/MS. Then they tried the HPLC (Alliance) but everything worked as expected. Intriguingly, if you changed the mobile phase fron 30/70 acetonitrile/water to 100% acetonitrile you got a nice peak; pity about not having any retention in the column. After quite a long time and much coming and going from Waters technical support they told us that the first quadrupole was dirty. They changed it and the signal was back to normal. Apparently they tried ultrasonic cleaning of that quadrupole to no avail.
Hope it helps

What was your application? If possible please provide type of samples (food matrix, blood, urine, water, etc...) and analytes (pesticides, proteins, peptides, etc...), thank you!

Yes the Xevo allows you to choose different resolutions (independently on both quadrupoles). You will lose some intensity in going down to 0.5Da, so if you are on the high resolution setting, that may be unhelpful. But the main problem is if you're on high resolution and the mass calibration drifts at all, in which case the signal may begin to drift significantly outside the window. Personally I haven't had the problem, but I remember it being mentioned during training. I see little point in the high-resolution setting anyway, because near-isobaric compounds need much much much better than 0.5Da if they're to be distinguished. I can't think of any two conventional compounds that could be distinguished at 0.5Da and not at 0.7Da!

I will look into this, thank you!

Thank you all for your support and guidance, I hope to figure this out soon enough. Will keep you all posted as progress is made. So far, no news is good news? System is still stable with good integrity, but overall response is still lower than normal.

[BHolmes]

I will look into this, thank you!

For our method, we have low and high resolution settings. For our current method; low end resolution is set at 2.8 and high end resolution is set at 15.1. I found these values in tune file and in calibration file. I hope these are correct locations for this information.

[carras]

Hello BHolmes,

sorry about this much delayed answer. We analyse residues of veterinary drugs in animal tissues and urine. With very few exceptions our methods use SPE to "clean" the samples.

[BHolmes]

UPDATE

Sorry for the delayed update, been working a lot on this problem and its still not resolved. But I have a fairly good idea what the problem is, the quads are dirty and need to be cleaned. This is the one solution that I has been suggested multiple times, on this forum and in discussions with other Xevo TQ-S operators.

Thank you again for all your tips and suggestions. I will post a final update after the FSE cleans the quads, I really hope this solves our problem!!

Brittany

[BHolmes]

LAST UPDATE FOR THIS POST.....ISSUE RESOLVED!!!

The Waters field service engineer ended up replacing the ENTIRE mass analyzer (MS1, Collision Cell, and MS2) for a brand new one - thank goodness for service contracts, that was a $40,000 part! There were some other problems (bad circuit boards, connections, cables) that surfaced during this replacement that were fixed also, so I can't be 100% sure it was the analyzer or a combination therof.

All in all, becuase of the larger sample cone for the Xevo TQ-S, a lot more matrix/sample gets put into the detector and it can build up over time resulting in charging between the quadrapoles and therefore reducing the overall signal. Now normally the quads don't get cleaned during the annual PM, but if you are injecting dirty samples (food matrices, biological, etc...) and injection volume is greater than 1uL I would suggest you have them cleaned at least annually if not more. In addition, I would suggest that the ion block should be cleaned at least 2-3 times per year.

Thanks again for your advice!
Brittany