LC-MS/MS sensitivity loss after few sample injections

[b95203035]

Hello everyone,

The LC-MS system I am using is ABsiex API3000 coupled with Water Acquity UPLC and the column is Hypercarb 3um 1.0*100mm. The problem I encountered recently is after running first three purified biological samples, the peak areas of standards dropped to only 50% and then after another three samples, the peak areas dropped to 30~40%, compared to the standard injected before running the samples. However, after overnight standby, the peak areas of standards return to almost the same, but the same problem appeared again after running the samples. I have tested 20 consecutive injections of standards only, the peak areas were stable. Could it be the ion source, tubing or column contamination by biological matrix?

The samples I injected were tissue extracts but purified by one LL extraction and two different SPE. And the LC method is first 6 min hold then 5% to 95% ACN gradient followed by two gradient washing step and finally 10 min equilibration.

Thank you in advance for your great help!

Yu

[sav]

It's classic matrix effect (ion suppression). You need to optimize washing step, but it can be very tricky with Hypercarb.

[b95203035]

It's classic matrix effect (ion suppression). You need to optimize washing step, but it can be very tricky with Hypercarb.

Dear Sav,

Thank you for your reply. Do you have any recommendation of the washing steps for Hypercarb column?

My LC method is
0-6 min 40uL/min 5%ACN (Hold)
6.1-15 min 40uL/min 5%-95%ACN (Elute)
15.1-19 min 120uL/min 95%ACN (Wash)
19.1-23 min 120uL/min 5%-95%ACN (Wash)
23.1-27 min 120uL/min 95%ACN (Wash)
27.1-35 min 120uL/min 5%ACN (Equilibrate)
5.1-40 min 40uL/min 5%ACN (Equilibrate)

Thank you so much!
Yu

[sav]

Do you have any recommendation of the washing steps for Hypercarb column?

Only experimental approach will work in this situation (information about target matrix compounds is absent). You can try these washing eluents (500 - 1000 ul; single washing step): MeOH, MeOH:MeCN (1:1), MeOH:iPrOH (1:1), iPrOH:MeCN (1:1), DCM:MeCN (DCM - dichloromethane; 1:1), MeOH:DCM:MeCN (1:1:1). Note: after eluents with DCM flush column with MeCN (500 - 1000 ul) before switch to gradient start point composition of eluent.

[b95203035]

Sav, thank you for your suggestion. However, my supervisor decided to insert two wash-injections after 15 min sample run. The injections were as follows:

1. 50%THF for 20 column volume
2. 3 cycles of gradient wash with MeCN/H2O
3. Matrix sample spiked with standard or Standard only solution

4. 50%THF....... repeat as above.

Unfortunately, there was still no improvement even though the retention time was consistent. The standard's peak area of first injection was high but keep dropping after each injections. And the weird thing was after let the machine standby overnight, on the 2nd day, the peak area of first injection was back to the same level as the first injection in previous day.

Now, I am not quite sure if the problem is on the hypercarb column. Should I invert the column and clean the column as manual's instruction? Or actually the ion source of our mass spectrometer was too dirty and needed to be cleaned?

I am at my wit's end..... please help me!

ps. my target is glutathione adduct in bovine retina

[sav]

The standard's peak area of first injection was high but keep dropping after each injections. And the weird thing was after let the machine standby overnight, on the 2nd day, the peak area of first injection was back to the same level as the first injection in previous day.

So, is this suppression appeared with pure standards?

Now, I am not quite sure if the problem is on the hypercarb column. Should I invert the column and clean the column as manual's instruction? Or actually the ion source of our mass spectrometer was too dirty and needed to be cleaned?

I think that you need to check "column version" little more. Try to replace the hypercarb column by restrictor (capillary tubing with id < 0.1 mm; adjust length to system pressure around 50-80 bar). Inject the standard (pure, without matrix) with 20-30 min lag (3 injection per point; isocratic condition).

ps. my target is glutathione adduct in bovine retina

I've successfully developed LC-MS method for glutathione in biological samples using HILIC mode (normal HILIC sorbent).

[James_Ball]

Those instruments can also have problems with charging of the quadrupoles when they become dirty. If they are dirty the sensitivity will drop off during analysis over time.

After sitting overnight, infuse a tuning solution and scan with your analytical method in manual tune mode and watch the TIC as it acquires. If the signal falls steadily over a 15-30 minute scanning time then you may have dirty quads.

[b95203035]

Hi Sav, James,

Thank you guys for your reply! We did some further tests this week, please allow me to describe our observation in detail:

1.When we inject the pure standard only, the response was highest in the first injection, and started to drop slowly in the following injections, but it would finally stabilize to some point.

2.After it stabilized, when we injected the sample matrix spiked with pure standard, the response of the standard keep dropping quickly: first -20%, second -30~40%, third -40~60%.....

3.After these matrix+standard injections, we again inject the pure standard only, the response was only around 1/3 comparing to the injections before sample matrix injections

4.If we let the API3000 standby for ~24hr, the response of the pure standard was back to ~70% of the initial condition and if standby for >48hr, the response was almost back to 100%.

5.We tried to change to another column with a largger i.d, but still the same problems again.

I did the manual tuning on our API3000 as James suggested and +Q1 TIC signal kept dropping in the 30min scan time while +Q3 was stable. Does this result mean that our quads are dirty? And the culprit of the sensitivity dropping issue is the dirty quads not the dirty hypercarb column?

Thank you!

[sav]

Does this result mean that our quads are dirty? And the culprit of the sensitivity dropping issue is the dirty quads not the dirty hypercarb column?

I think its suppression of ionization by matrix compounds (hypercarb is very retentive sorbent; p. 2) and first quad (Q1) troubles (contamination or/and electronics; p. 4 and James method)

[James_Ball]

Have you had someone do a "rail clean" on the quads before? If not, then it is probably a dirty Q1 causing some of the problem, especially since it also happens with syringe infusion without the column.

The Q0 you can clean yourself easily. It is the four curved pieces that you see if you remove the heated interface and the cone shaped piece just inside. You can use a long swab and some methanol to give those a quick clean and see if that helps. If it is deeper in the instrument it means pulling the analyzer out and that isnt so easy.

[b95203035]

Hi Sav, James,

Thank you for helping me analyze the problems!

Our last PM was in June, I believed they cleaned the quads at that time. And we have cleaned Q0 ourselves, but no luck. Probably we should ask them to clean the quads even though it has just been five months since last PM.

So, in your opinion, the dirty quads might be the main cause of our situation: stable during continuous standard injections --> irreversible loss of sensitivity during and after continuous matrix injections in the same day --> Sensitivity back to normal after instrument standby for >1 day?

I don't know if we could rule out the hypercarb problem or not. If unfortunately the problem is on the column, we really have no idea how to deal with it since we'd already tried THF as washing solvent but still no improvement.....

Thank you!

[James_Ball]

I have never used the Hypercarb column you listed, but would it clean up after sitting filled with mobile phase overnight? I would think a column problem would clear up better with flushing rather than sitting idle, that is why I was thinking possibly dirty quads causing charging.

[sav]

So, in your opinion, the dirty quads might be the main cause of our situation: stable during continuous standard injections --> irreversible loss of sensitivity during and after continuous matrix injections in the same day --> Sensitivity back to normal after instrument standby for >1 day?

May be, but, i can't denied "column effect".

[carlo.annaratone]

Have you tried running the MS in full scan mode, to see if there are signs of background rising, or contaminants peaks?

[cerasinol]

Have you solved the problem of loss of sensitivity during a sequence of analysis?
I have the same problem.