Problem on thermo lcq

[mmegliol]

Hello,

I have a little problem (I hope!) I tried to run the calibration on my lcq but during the process the step zoom scan failed, any idea what could be!?

Is it possible my calibration solution is too old?

Thanks,

Matteo.

[lmh]

zoom scan is always the one that fails if anything fails (OK, sometimes trapping fails too). It's OK to recalibrate doing only zoom-scan (i.e. don't repeat the whole thing). Your calibration solution should be fairly fresh (I wouldn't recommend storing more than a few weeks) because once it's got acetic acid in it, the Ultramark tends to decompose and you get a parallel series of peaks about 10m/z less than the correct peaks. On a low-res ion trap there's a real risk the calibration will go for the wrong peak! What I would suggest, is if you're starting by tuning to caffeine as used to be recommended, try tuning to MRFA before you calibrate (I think Thermo recommend this now anyway). Sometimes the calibration fails because it can't find enough of the high masses if it's tuned to the lowest mass. MRFA is a better compromise.

[chemex]

Hello, I can only tell you to prepare a fresh calibration solution. It is suppose to last one month but mine always goes off after a couple of weeks.

While I am here, my company bought my mass spec (lcd deca xp plus) second hand and I am missing the LC Tee Union. I use this to tune the MS with a specific analyte. I start to received quotes for this part and it is quite expensive. My boss thinks we can manage without but I don't. I need a tune method for every signe compound, don't I?

[chemex]

Dear All

I manage to buy a HPLC Tee connector. The issue now is that the LC flow is high (200-400ul/min), I infused my analyte at a rate of 5ul/min with the syringe. Both flow are supposed to go in the MS but the LC flow splits between the MS and the syringe. I can see the liquid dropping from the end of the syringe. How have managed this, any advice will be more than welcome. I am not sure that the compound is actually going in the MS

Another question is that, I infuse glyphosate directly in the MS but the molecular ions or fragments are not detected (full scan mode). Any ideas why?

Many thanks

[MaryCarson]

It sounds like either 1) your syringe needle fitting is too loose; 2) the back pressure on your MS is too high. It is easy enough to check the MS back pressure--look at the pressure on the LC while mobile phase is going through column into MS, then loosen the connection to the MS. If the pressure drops by more than 10% or so you might be looking at a blocked or partially blocked capillary. Depending on the model, you might be able to sonicate the capillary to remove the blockage.

As to fixing the syringe...there exists this nifty little stainless steel doodad needle-thingy [that's sophisticated techno-speak] with a luer-lock on one end and a male LC tubing fitting on the other end. I have no idea what it is really called or where to buy it (ours came with one of our LCs...), but it sure is a lot more convenient than trying to compression fit a gas-tight syringe needle into what you hope is the right size tubing into a ZDV LC connector.

[chemex]

Mary

Thank you for your answer. The problem is resolve for now. I change the tube lenght and ID, I also "play" with the introduction angle of the sample. It seems to work even if you have to check it during every single infusion for leaks. The syringue fitting needle is indeed loose.

I infuse directly HPLC grade water and the MS is detecting multiple m/z. In fact, the solvent noise is soo high that even a 1ppm of an analyte in water/MeoH 50/50 (direct infusion) is not "seen" by the MS.

I try two differents hplc-type water, different clean syringues and different tubings. The instrument is second hand so the tubing is not new. I might think about buying a new one.

I have the same result with an injection of a concentrated sample from the autosampler, which is "shadowed" by the mobile phase noise and therefore not detected.

Any ideas of what I should do to get rid of the solvent noise?

Many thanks

[lmh]

You will always get a leak back out of the syringe if you get "normal" Hamilton-style syringes with a simple metal plunger. You absolutely must, must, must get gas-tight ones (recognisable by a white plastic tip on the end of their plunger).

The reason I say this so strongly is that if you are teeing together 300uL/min from the LC pump and 5uL/min from the syringe, and combining these to flow into the spray chamber, then even the tiniest leak in the syringe will mean that Absolutely No analyte will get to the mass spec at all! You only need a back-leak of 5uL/min, which is barely noticeable, to give you no signal except solvent. Even the gas-tight syringes do not last for ever. When they get loose, it's time to replace them, and false economy to try to make do with an old leaky one.

What solvent signal are you seeing? If you have specific ions, like 279 and 391 in positive mode, then you do have dirty solvents or tubing. If it is a general fuzzy mess through the whole range of the spectrum, there may be nothing wrong. If you have a ladder of peaks, it might be a salt contamination from somewhere, but I'd be very surprised at this.

If you're getting no signal when you do an injection from the autosampler in the absence of any column, then there are a few other things to check. If this is a Surveyor autosampler then it has a double-diameter syringe thing with a tiny inner plunger (which makes the injection) and a wider outer plunger (which does the high-volume washes). The only thing that decides whether the motor drives the inner, narrow plunger, or the outer, wide plunger, is how far up or down it is. If there is a misunderstanding between the software and the syringe about how far up or down it is, then it will be using the wrong plunger, and it may squirt the sample right through the sample loop, and to waste. Always, Always before you start a sequence, purge the syringe, make sure it's full of liquid (a few small bubbles are acceptable if they stay the same through purging), and then home it (in the direct control menu of Xcalibur). If it makes grinding noises, it was in the wrong place. If this happens often, clean the lead-screw that drives the syringe, and check the needle isn't blocked.
Also do check that the tubing from injection port to injection valve is original and hasn't been cut/shortened/modified; its volume is critical and should match a value in the Configuration set-up.
You can also check the needle isn't blocked (especially if the syringe makes grindy noises, see above). Use Direct Control to get the needle to its removal position, locate the little black plastic bit at the top of the needle and turn sideways to unlock, and draw it out upwards, carefully. Now do a needle wash from direct control and make sure the liquid comes out as a free steady stream. Remember when you do this that the instrument will move the needle position to the wash-station, so be prepared for it to whizz around. If you've freed the needle tubing you will be OK! To do this, you may also need to disable the bit in configuration about sensing whether the door is closed, or fool it into thinking the door is closed by putting a small magnet at the top - but Be Careful not to get injured by moving parts. Do this at own risk, with sense! Good luck!

[chemex]

Hello lmh

Thank you for your answer. It is true that with a normal Hamilton syringe, I always have to check any single time if they is no back-leak event without the LC flow. I wonder how other users of LCQs manage. I will definitely look for those gas-tight syringes. It will save time and at least I won't have doubt whether the sample is actually going into the MS.

For the noise problem, the problem is that i don't see specific ions. Apart from the 181, all the others come randomly. I inject hplc water and a mixture water-meoh. The noise still there. When I inject the analyte, I have the same profile because of this noise.

I try to tune and calibrate the instrument. For those who are familiar with the instrument. My normalisation level (NL on Tune plus) is around 1e+003-1e+002, its suppose to be higher than that (1e+007 or am I wrong?). The ionisation time is 200ms but suppose to be <1ms. I have set the max injection time at 200ms like the manual said. I can change it manually to <1ms but I loose the calibration ions. Does anyone know how to increase the NL and deal with the IT.

I must add that the AGC (Automatic Gain Control) is set to 5e+007, full MS target.

Thank you and enjoy your weekend.

[lmh]

I use a LCQ DecaXPplus, and I'm not sure how much this varies between models.

Yes, your signal sounds far, far too small. I would expect about 10^7. 200ms sounds like the automatic gain has hit its limit. What actually happens in each scan is that the entrance lens is opened for a very short time, and then the trap is scanned extremely fast (with dreadful resolution), while the instrument simply sums the total number of ions. You don't see the spectrum from this scan. It then calculates how long the entrance lens needs to be open to fill the trap optimally. The result of this calculation is the ion time, and is then used for the proper scan, whose data you see. If the calculated ion time is more than a certain value (in your case I suspect 200msec), the entrance lens is opened that long anyway, even though it means you will not have enough signal. The reason is that if the solvent is nice and clean and no peak is eluting, the instrument might find it needs 15 seconds ion time, and a whole peak could come and go during that time....

Does your signal disappear completely when you are not putting any flow into the spray chamber at all? It should vanish. Putting the solvent in should make a much larger noisy signal than no solvent whatsoever. In the situation you are in, I would try infusing calibrant directly from the syringe, without any flow from the LC, at a sensible flow (5uL/min?), and see if you can see the calibrant masses. If you can't, then something is seriously wrong and there's no point in progressing to real samples and standards. If you do see calibrant masses, then I'd suggest tuning to MRFA, then using this tune-file, infusing your test chemical and seeing if it's now visible. Incidentally, for calibration, it's much better to tune to MRFA than to caffeine before you calibrate.

[chemex]

Hi lmh

I am glad to read your answer. I have try few things that you suggested.
I have increased the AGC to 1+008, I can see the calibration ion except the 195. When the max inj time is set to 200ms the ion are present but the spectrum is noisy. I have set the IT to 0.99 and the spectrum is less noisy. However the NL is better (1+003-1+004) but still far from 1+007.

I have tuned the inst with MRFA. I tried the calibration but most of it failed mainly because of the signal of 195 which is not high enough. All the other ions were found. I have prepared another calibration solution with 400ul of caffeine but it's still not detected. I will try to call thermo even if we don't have a contract with them. I don't know what to do now to deal with the NL

Many thanks

[chemex]

I forgot to say that I'm infusing the calibrant directly and yes when i stop the infusion, there is no peak on the spectrum after few seconds.

[lmh]

It sounds like there is something seriously wrong with ion-transmission in your instrument. You should get a noisy signal of at least e+06 when infusing pure solvent, and I would be surprised if the instrument hit its maximum ion time even on very clean solvent. When infusing calibrant it should need only a very short ion-time and you shouldn't need to alter the ion-time manually as the automatic gain control, if aiming for anything in the region of 1*10e07 to 1*10e08 should be filling the trap completely with calibrant ions in not more than a couple of ms.

The fact that the signal disappears when you turn the syringe off is important. It tells you that the needle isn't completely blocked, and nor is the ion transfer capillary.

The fact that most of your ions are not getting into the instrument suggests a more serious problem. There are lots of possibilities, unfortunately, so I think you are right to contact Thermo. One possibility is that somewhere an electrical contact isn't being made, or has been shorted. If something is shorted, you will see a very low value for the appropriate voltage in the information pane in TunePlus (the pane that shows all the voltages, temperatures, gas-flows and turbo speed etc., that is displayed optionally on the right hand side of the screen if you select its button, the button with a picture of tiny rows of text on it). If something isn't actually connected, then there will be no voltage on the actual lens (or whatever), but the read-back will be fine. This happens, for example, if someone has assembled your ESI probe incorrectly. It might, next time you vent, be worth doing a quick check that there isn't vast amounts of dirt visible on the API stack components (bit behind the spray-chamber, that you remove when vented), and that the tube-lens and skimmer were plugged in properly and connected to their electrical cables last time someone cleaned it. The tube-lens, for example, guides the ions round an S-bend out of the ion transfer capillary and onto the skimmer, before the end of the quadrupole ion guide. If it is left unconnected, the ions will presumably go in a straight line and land up stuck on the side of the skimmer's cone.

By the way, I'm assuming your instrument actually has some voltage on its ion detection system. The multiplier voltage is probably about 1200V?

In positive mode, you are actually aiming for a very noisy signal at about 1e+07 on pure solvent; that would be ideal. In negative mode there will be less signal and less noise.

[chemex]

Hi lmh

I now think like you that the problem is more serious that I thought. I just hang up with one engineer from thermo. According to him, the thing to look at is if the capillary is not blocked. I pull it out and it is being flushed with solvent right now (it wasn't blocked at all). in case it don't change anything.

In the mean time, I have check the information panel. these are the reading while running.

ESI Source
|Spray Voltage (kV)|: 4.86
Spray Current (µA):-0.00 (range between 0 to 100uA during the run)
Sheath Gas Flow Rate: 49.76
Aux/Sweep Gas Flow Rate: -0.16
Capillary RTD OK: Yes
Capillary Voltage (V): 142.42 (set at 41V)
Capillary Temp (°C): 249.50
Tube Lens (V, sp): 55.00

Vacuum
Vacuum OK: Yes
Ion Gauge Pressure OK: Yes
Ion Gauge: On
Ion Gauge (x10e-5 Torr): 2.41
Convectron pressure OK: Yes
Convectron Gauge (Torr): 1.38

Turbo Pump
Status: Running
Life (hours): 37827
Speed (rpm): 45000
Power (Watts): 86
Temperature (°C): 54.00

Ion Optics
Multipole Frequency On: Yes
Multipole 1 Offset (V): 101.95
Lens Voltage (V): -81.46
Multipole 2 Offset (V): 102.46
Multipole RF Amp (Vp-p, sp): 400.00
Entrance Lens Voltage (V): 102.97
Coarse Trap DC Offset (V): 102.75
Fine Trap DC Offset (V): 20.00
Analyzer Temperature (°C): 29.74

Main RF
Reference Sine Wave OK: Yes
Standing Wave Ratio OK: Yes
Main RF DAC (steps): -97.00
Main RF Detected (V): -0.01
RF Detector Temp (°C): 39.94
Main RF Modulation (V): 0.02
Main RF Amplifier (Vp-p): 8.02
RF Generator Temp (°C): 36.82

Ion Detection System
Conversion Dynode: On
Dynode Voltage (kV): -14.90
Multiplier Voltage (V): -1776.64 (this value is really low, isn't it?)

Power Supplies
+5V Supply Voltage (V): 5.07
-15V Supply Voltage (V): -14.99
+15V Supply Voltage (V): 14.82
+24V Supply Voltage (V): 23.63
-28V Supply Voltage (V): -28.15
+28V Supply Voltage (V): 28.33
+28V Supply Current (Amps): 1.05
+35V Supply Voltage (V): 35.59
+36V Supply Voltage (V): 35.71
-150V Supply Voltage (V): -149.27
+150V Supply Voltage (V): 150.71
-205V Supply Voltage (V): -212.25
+205V Supply Voltage (V): 207.80
Ambient Temp (°C): 31.15

Instrument Status
Instrument: On
Analysis: Idle
Retention Time (min): 0.00

Analog Inputs
External channel 1 (V): -0.00
External channel 2 (V): 0.00

Syringe Pump
Status: Running
Flow Rate (µl/min): 5.00
Infused Volume (µl): 10.00
Syringe Diameter (mm): 2.30

Digital Inputs
READY IN is active: No
START IN is active: No
Divert/Inject valve: Load

I seriously thinking about calling an engineer for at least one day to set everything and I can at least start running samples. Thank you for your suggestions. You have been a massive help.

[lmh]

Good luck, it sounds like Thermo are being helpful.
Yes, the spray current looks low. Mine is about 0.35uA when infusing 5uL/min 50% MeOH. See comments on spray voltage at the end of this post!

I'm not sure I understand the capillary voltage setting; mine cannot be set that high on a DecaXPplus, but maybe your model is different.
No, that's not too low a multiplier voltage; if anything I'd consider it high on this instrument. My instrument is currently 1278V. I think Thermo prefer not to push their EM horns to the limit.
The current on the 28V supply matches mine.

I assume the capillary that the engineer is getting you to clean is the ion transfer capillary that you unscrew with the little metal tool and pull out from the API stack, that sticks straight into the instrument. Good idea to check that! If you haven't ever done so, and feel confident to give it a go, you might have nothing to lose by venting the instrument, taking out the API stack, and checking it all looks OK. This is a user-serviceable part and there should be instructions in the manual - it's basically undoing the electrical connector on the right hand side of the stack, and pulling the whole thing out forwards towards you (it's only held in by the vacuum). That and the capillary are about the only bits that a user can really do much with.

(**) The other thing that would produce the symptoms you describe is if you have absolutely no spray voltage. I know you do have one on screen, but it's just possible that if something's seriously wrong with your ESI probe (someone's reassembled it with a bit missing, so the contact isn't there, or the cable that supplies it with high voltage is damaged or incorrectly plugged in), then no volts will be reaching the spray needle. I'm not sure how to test this, beyond the highly risky approach of applying 4kV in negative mode and edging the probe position closer and closer towards the sample cone/end of capillary until you begin to see a faint glow of corona discharge. If you get one, you definitely have volts! If you don't, there's something up! But please be careful if you try this; if you actually get sparks from needle to sample cone you may kill bits of electronics.

[chemex]

Hi lmh

The spray current is unstable during all the run. Goes from 0 to 100uA. The only reason I found is that I might have a problematic mixture of solvent. But I'm infusing a mixture of methanol and water. May be I should just try methanol only.

The capillary voltage is indeed high. It is set at 41V but is always around 140V. I can't find any reason why.

I cleaned the ion transfer tube. Yes the one you unscrew with the metallic tool. Flushed it with solvent. Didn't seem blocked. Try an infusin this morning, nothing has changed.

Like you suggest, I am going to pull out the API stack and clean the lens and skimmer. I am trying to convinced my boss to have an engineer to service the instrument.

I will think about the "spray voltage test". I am sure that from time to time (when I check the source during the run), I can see a blue discharge between the end of ESI needle and the ion sweep cone.

Many many thanks. I not really confident about completely switching off the MS and inspection the API stack. I will let you know if this improve my signal intensity.

Thank yoiu