Chromatography Forum

Seeing is believing, but I still find it inexplicable.

Let's review:

1. You inject 1 ul of a 1mg/l solution splitless, that gives 1 ng per analyte on column. This give speaks that are too small for your purpose, and much smaller than you can achive injecting the same solution onto a different make of instrument.

2. As well as the peaks being too small the background at the retention times of the peaks is too high. The way that the background declines, and the changes in how it declines look like the tail of a solvent peak.

3. The MS detector passes its detectability tests with a different analyte and solvent.

4. The MS and the inlet consumables have been cleaned and / or swapped repeatedly, with no improvement in performance.

Troubleshooting:

Can you run the same phthallate solution on another instrument, even an FID ?

If you run a 100 mg/l solution what do the peaks and solvent tail look like ?

Do the spectra of the peaks confirm that they are what they are supposed to be ?

Does the spectrum of the background confirm that it is toluene ?

If you run any other solution on the GC-MS; alkanes in hexane for example (but not the Varian test mix), are there the same problems ?.

Peter

Troubleshooting:

Can you run the same phthallate solution on another instrument, even an FID ?

If you run a 100 mg/l solution what do the peaks and solvent tail look like ?

Do the spectra of the peaks confirm that they are what they are supposed to be ?

Does the spectrum of the background confirm that it is toluene ?

If you run any other solution on the GC-MS; alkanes in hexane for example (but not the Varian test mix), are there the same problems ?.

Peter

I appreciate your help, however, without anyone having the same GCMS and experience with it there'll be no progress. Again, I'm really not a novice neither to GC nor to MS. I've made these tests more than once, spectra confirm it all, it runs on other machines (including FID, but not Varian), the summary is always the same: low sensitivity and extreme solvent tailing in splitless mode.

The easiest solution would be another lab here on the forum with exactly this instrument who can tell me either

- we use splitless injection for PAH in the low ppm range with success
or
- we tried and the instrument is not able to do it

It's not meant to put Varian down - we have lots of Varian spectrometers in our lab without any problems.

I've made these tests more than once, spectra confirm it all, it runs on other machines (including FID, but not Varian)

It would have been useful to have known that earlier.

Peter

It would have been useful to have known that earlier.

See my very first post

rp18,

1) You inject toluene at 90C with a 0.25 id column that I assume is 30 meters or so long and you get your solvent peak at 5 minutes? That is a really long elution time for 1 to 1.5 mL's/min. 2) You only have about 5 minutes of retention time separating the first of your phthalates from the solvent front. 3) You have your filaments on while your solvent is eluting. If you could increase the separation between the solvent and the first peak you could increase your filament off time to decrease the impact of the solvent on the filaments...

Just out of curiosity, can you post an OFN chromatogram and conditions, as well as what solvent that solution is in? Can you also post the spectra of the first bigger peak at 11 minutes?

1) there's a solvent delay of 5 min. Peak is much bigger before. Solvent is eluting until the end of the chromatographic run.
2) Phthalates elute quite late, gradient is part of an upcoming ISO norm, and again, working well on several other instruments that passed a ring test.
3) No - filament off for 5 min - that's a time USUALLY long enough the get all solvent out - except for THIS instrument type - and that's the main problem I think.

OFN doesn't show this simply because the masses of the solvent are below the scan range of the method...
Sure I could now modify the phthalate method and start scanning at m/z 100. But that's not an option for me.

To summarize: my suggestion is that the injector has a dead end somewhere in the split line (by dirt OR by design), so solvent and compounds go there and elute slowly. All I want to know is if there's anyone having similar problems.

I do not analyze the type of compounds you do but I have attempted to use toluene in the past. I have used it on traps and quads and found a similar response as yours. A large tailing solvent peak that lasts long after the splitless injection time. I then see toluene ions (91) all along the baseline interfering with my target peak spectra. The baseline is high increasing my S/N. In the trap that situation will reduce your signal for your target compounds that are in lesser amounts. Toluene always seems to contaminate my vent lines and septa after using it. My 5973's react the worse but the signal is more suppressed on the traps.
Can you use another solvent? I have used ethyl acetate, methanol, and hexane and not experienced this issue. If you have to use toluene you might try a higher starting temperature even though it is a splitless injection. Try injecting just solvents in your program and see which has this drawn out solvent front.
Seaox

Seaox,

thanks a lot! At least I'm not alone with this.

I've tried other solvents (hexane, octane, cyclohexane - the usual ones) and they all show this more or less, with toluene being by far the worst.

Funny thing is, my 5891 (and the other GC's of the ring test) don't have this problem. No 91, flat baseline.

We'll keep on searching for a solution.

I'm still wondering what it is about toluene the makes it a problem . The aromatic ring is an obvious candidate, so what happens with benzene and xylenes

Peter

I have the same problem in Varian ion trap, and I do PBB and PBDE analyse. and use the tolune as solvent.
To rp18,
I know you have tried " other solvents (hexane, octane, cyclohexane - the usual ones) and they all show this more or less, with toluene being by far the worst. "
Do you really use the Varian suggested splitless injection mode?
Initial on 20:1
0.00 off off
0.7 on 100:1
5 on 20:1
This is really important because compare to SQ, which is like a filter, IT is more like a beaker, when too many tolune ion gather in ion trap, it can not exclude easily.
I really suggest you to change the split pencil filter, maybe too many tolune assemble there, which can escape to the 1177 injector or lead to ponding of the split way.
At last, i suggest you change the solvent, and narrow the mass range 95-300, if possible. of cause, you can get nice sensitivity with the help of MS/MS, not SIS. because MS/MS is the advantage of ion trap.

To rp18,
Do you really use the Varian suggested splitless injection mode?
Initial on 20:1
0.00 off off
0.7 on 100:1
5 on 20:1

Yes. And tried several others.

I really suggest you to change the split pencil filter, maybe too many tolune assemble there, which can escape to the 1177 injector or lead to ponding of the split way.

Filter was changed already.

At last, i suggest you change the solvent, and narrow the mass range 95-300, if possible. of cause, you can get nice sensitivity with the help of MS/MS, not SIS. because MS/MS is the advantage of ion trap.

It would be interesting to know if this is a general trap problem and not - as we suggested - an injector problem. We don't (yet) have MS/MS.

Currently the Varian is off and the good ole HP GCMS does the job.

It would be interesting to know if this is a general trap problem and not - as we suggested - an injector problem. We don't (yet) have MS/MS.

Currently the Varian is off and the good ole HP GCMS does the job.[/quote]

One of my friend use Saturn2200, which is the old style model of 220-MS, and she bought it at least five years ago, but it works well now. She said to phthalates like BBP、DBP、DEHP、DNOP, which has one peak, even 0.1ppm standard can get a nice peak with Saturn 2200, and to DIDP、DINP, which has many peaks, about 5ppm. She found there is no difference between Saturn 2200 and 5973. She use it to determine the trace phthalate esters in plastic everyday.
But I am not sure about what's your problem. It seems it's not common. Trust me, I know what's your feeling, and I have some reference papers in ion trap in phthalate if you need, i could email to you.
Another question，Which kind of EFC in your 1177, type 1 or type 5?

EFC is type 21

It would be nice to get some method details (and the papers), especially injector type and settings!

Thanks a lot.

GC Conditions
Column: FactorFourâ„¢ VF-5ms capillary column,
30 m x 0.25 mm x 0.25 μm
(Varian Part Number CP8944)
Inlet Temperature: 280
Injection Volume: 1 μL
Carrier Gas Flow: Helium at 1 mL/min
Injection Mode: Splitless, pulse injection at 35 psi
for 0.6 min
Oven Program: 50 oC for 2 min, to 280 oC at 30 oC/min,
to 310 oC at 15 oC/min and hold for
3.33 min, for a total run time of 15 min
Washing Solvent: Cyclohexane
MS Conditions
Filament Delay: 5 min
Ion Trap Temp: 220 oC
Manifold Temp: 50 oC
Transfer Line Temp: 280 oC

Compound Name Quantitation Ion
Diethyl phthalate 177+149
Benzyl Benzoate (IS) 105
Dibutyl Phthalate TIC
Benzyl Butyl Phthalate 206+149
Di-(2-ethylhexyl) phthalate 279
Di-n-octyl phthalate 279
Diisononyl phthalate (DINP) 293
Diisodecyl phthalate (DIDP) 307

Thanks!

PS: is it possible to post/send me a chromatogram?