625 & Varian 3800

[Bsegokc]

Problem: Pentachlorophenol and Benzidine response at 5 to 10% of DFTPP/DDT and tailing is above method spec. Loss of spectral integrity beyond 50ng on column for DFTPP and non-active compounds.

System

Varian 3800 GC w/ 2200 Ion-trap MS
1177 split/splitless injection port
RXI 5SIL-MS with 10m integra guard 30m x 0.25 x 0.25
4mm ID single-gooseneck Restek Sky Liners; necked end @ column end

Run program:

Oven start: 40C
Oven ramp: 40C 1minute to 340C @ 10C/min
Injector: 250C
Flow: 1ml/min Ultra UHP Helium w/ clean background spectrum
Split: On @ 0.15min 50:1

Injection volume: 2ul

Column, liners, nuts/ferrules and injection port are new. Column is low bleed. No field samples have been processed. So I'm thinking method optimization. Anyone have experience with these Varian GCs and running SVOC by ion-trap?

Thanks!

[Hornet]

Problem: Pentachlorophenol and Benzidine response at 5 to 10% of DFTPP/DDT and tailing is above method spec. Loss of spectral integrity beyond 50ng on column for DFTPP and non-active compounds.

System

Varian 3800 GC w/ 2200 Ion-trap MS
1177 split/splitless injection port
RXI 5SIL-MS with 10m integra guard 30m x 0.25 x 0.25
4mm ID single-gooseneck Restek Sky Liners; necked end @ column end

Run program:

Oven start: 40C
Oven ramp: 40C 1minute to 340C @ 10C/min
Injector: 250C
Flow: 1ml/min Ultra UHP Helium w/ clean background spectrum
Split: On @ 0.15min 50:1

Injection volume: 2ul

Column, liners, nuts/ferrules and injection port are new. Column is low bleed. No field samples have been processed. So I'm thinking method optimization. Anyone have experience with these Varian GCs and running SVOC by ion-trap?

Thanks!

I have similar problems with PAH analysis, system performance suddenly drop after i run several samples and heavier PAH disappear and come back on my chromatogram, i'm working for 3 months with PAHs and still can't get a method stability. My opinion (at least until now and based on my experience) is that this ion trap isn't the best option for very high boiling compounds because it gets dirty very easily, and the trap can't reach high temperature (unlike other mass spectrometers, i/e quad.) so you have to clean it manually very often, depending on your samples load. Keep on mind that i have to try many other things yet, so my judgment can still change.
Your case seems different, you said you didn't run samples so it doesn't seem a trap contamination problem (but you never know).
Answer these questions and maybe i can help you better:
- What kind of semivolatiles are you analyzing (or want to)?
- What are you Xferline and trap temperature?
- Do you have silchrom or normal electrodes? (silchrom are blue ones)
- Your liners are top notch, but you keep the gooseneck at the bottom or on top of the injector? And did you try your system performance test after changing the liner with a new one?
- What do you mean with "Loss of spectral integrity beyond 50ng on column for DFTPP and non-active compounds."? Your mass spec. seems very different from the table 3 of EPA 8270 (or table 9 of epa 625)?

I hope i can help you and if anyone else runs high boiling semivolatiles daily on this trap, please tell us your experience.

[Bigbear]

On first glance I would say you are injecting too much. You may be overloading the injector. I run 525 using a pulsed splitless injection of 1ul @ 50 psi.

[Hornet]

On first glance I would say you are injecting too much. You may be overloading the injector. I run 525 using a pulsed splitless injection of 1ul @ 50 psi.

Depends on the solvent he's using. With methylene chloride a 1177 single gooseneck splitless liner can almost handle 2 microliters without the need of pressure pulse, but is fine with hexane for example. With your injector conditions (assuming a T of 250+ °C) you could easily inject 4 to 5 microliters.

I don't think his problem is injector overload, i have similar problems and i run a pulsed splitless at 40 psi.

[Bsegokc]

We've tried various flows, injection volumes, etc. Degradation problems are fairly consistent. I've run SVOCs for several years but using Agilent and Quads so there are little peculiarities.

For instance, 50ng of DFTPP on column and the 440 ions are 10 to 20 times higher than base peak. 5ng DFTPP and spectrum passes. The mass spectrum changes quite a bit for most compounds beyond the mid calibrator; I suspect interference in the ion-trap.

We're running the standard 625 list and my Agilent playbook doesn't seem to apply =) The mass spec is very sensitive but the working range degrades at around 50ng for most compounds when I'm required to spike 100.

You running with the gooseneck near the septum?

[Hornet]

We've tried various flows, injection volumes, etc. Degradation problems are fairly consistent. I've run SVOCs for several years but using Agilent and Quads so there are little peculiarities.

For instance, 50ng of DFTPP on column and the 440 ions are 10 to 20 times higher than base peak. 5ng DFTPP and spectrum passes. The mass spectrum changes quite a bit for most compounds beyond the mid calibrator; I suspect interference in the ion-trap.

We're running the standard 625 list and my Agilent playbook doesn't seem to apply =) The mass spec is very sensitive but the working range degrades at around 50ng for most compounds when I'm required to spike 100.

You running with the gooseneck near the septum?

Regarding the DFTPP spectrum, what you said seems normal to me, i mean 10 to 20 times higher seems a bit much (did you use baseline subtraction to get these values?) but the 442 m/z can be the base peak, and will probably be on this ion trap. Table 3 of EPA8270 confirms this.
The gooseneck should be at the bottom of the injector, to prevent contact between analytes and cold/active metal parts of the injector bottom, which will result in loss of performance.

Can you please answer the questions above? It's important to know Xferline and trap temperatures, electrodes type and the last time they were cleaned. Also it would be useful to know which SVOC are you analyzing and in which solvent.

[Bsegokc]

Baseline is subtracted and it's thankfully not picking up a point on the peak. 442 tends to be base peak at any concentration level. After 10ng on column it is about the only ion you see in the spectrum.

Ion-trap temp of 150
Manifold temp of 120
Transferline temp of 150

Solvent MeCl2

Refer to EPA 625 for our list of analytes

[Bigbear]

If you are using any type of autotune the 442 will be the base peak as those tunes stress the high end. For 525 you have the option of 198 or 442 as base peak.
I modify my autotune by letting it autotune then lowering the ion focuss so the 442/198 ratio is ok ( I use 442 as peak).

[Hornet]

Baseline is subtracted and it's thankfully not picking up a point on the peak. 442 tends to be base peak at any concentration level. After 10ng on column it is about the only ion you see in the spectrum.

Ion-trap temp of 150
Manifold temp of 120
Transferline temp of 150

Solvent MeCl2

Refer to EPA 625 for our list of analytes

Your peaks are tailing because your spectrometer temperatures are way too low for the analytes on that list. Remember that the transferline temperature must be at least as high as the highest oven temperature reached by the method (in your case 340°C), it's better if you set it at 20°C more if the column specs permit it.
Your trap temperature is too low as well, i would try to increase it at least to 200°C and see if it works better, you can raise it to 240°C without any problem for your instrument (the only problem would be low sensitivity for low-boiling compunds but if the instrument is dedicated this isn't a issue), and remember that at 250°C (highest temp. permitted) the silchrom on the electrodes will start to deteriorate.
The manifold doesn't require to be that high, default temp for analysis is 50°C but if you raise the trap at 200°C+ it will never reach 50°C, so you better put it at 70°C like i did and end of the story.

2 microl of methylene chloride on a 1177 splitless liner will overload your injector (from FlowCalc) if you don't put any additional pressure, resulting in loss of analytes/sensitivity, backflash, etc. 30 psi will be more than enough for a pulsed splitless pressure with that inj. volume.

So: try higher spectrometer and transferline temperatures, and try a pulsed splitless at 30+ psi.

IMPORTANT: i don't know how many injections have you made with these instrumental conditions, in my opinion the last 10-20 cm of your column end (xferline end) are heavily contaminated with analytes that didn't make it through the xferline due to low temperature. Before trying any of the above solution you must turn off the instrument and cut the column to prevent trap contamination.
If you still have resolution and tailing problem after all this operations your trap is probably contaminated because of the low temperature that overloaded it with heavy compunds, at that point you have to decide (basing your opinion on the number of injections you did) if a bakeout is enough or if you have to manually clean the electroes. Since you said you only injected standards and no field samples i think a bakeout should be enough.

Good luck.