Problem with detecting last 3 PAHs - SOLVED

[wraymogg]

Hello,

I have trouble detecting the last 3 PAHs:
- Indeno[1,2,3-cd]pyrene
- Dibenzo[a,h]anthracene
- Benzo[ghi]perylene

My settings are:
GC-MS: ThermoQuest GCQ Plus - ion trap quadrupole
Column name: Phenomenex Zebron ZB-5MS
Composition: 5 % Phenyl 95 % Dimethylpolysiloxane
Size: 30m x 0.25mm x 0.25µm
Ramp: 40 °C - 4 min, 240 °C - 25 °/min, 310 °C - 5 °/min, 310 °C - 10 min
Injector temp: 300 °C
Columns flow: He, 40 cm/sec
Injection type: Splitless
Needle rest in injector: 3s
Injection volume: 1µl (tried with 2 µl and 3 µl)
Transfer line temp: 300 °C
Ionization chamber temp: 240 °C
Standard: MRC with 1 mg/l (1000 ppb)

Tried both full scan and SIM, for SIM i used 4 time windows, with the ions for every PAH (Parent ion, Quantization ion, Qualifying ion)

Starting with Naphtalene and ending with Crysene everything is beautiful. Big peaks, good shape, good separation.

Benzofluoranthene elutes together with Benzo[k]fluoranthene in a very ugly shape, broadened. Same ugly, broadened and noisy form for Benzo[a]pyrene. But still identifiable and quantifiable.

However Indeno[1,2,3-cd]pyrene, Dibenzo[a,h]anthracene and Benzo[ghi]perylene are nowhere to be found. I tried A LOT of settings, with multiplicator, AGC, waveform filtering, scans per second. No luck.

I shortened the column, changed septa, liner, with new ones... nothing. Every other peak is very beautiful and strong S/N wise. Not the last PAHs...
Calibration is flawless, good signal, multiplier is auto set on 1300. Pressure in system is at 40mTorr, no leaks, no contamination.

Is it possible to be a column fault ? If not, what can it be ? Any advice would be appreciated.

Best regards,
Vlad Popovici

[James_Ball]

Hello,

I have trouble detecting the last 3 PAHs:
- Indeno[1,2,3-cd]pyrene
- Dibenzo[a,h]anthracene
- Benzo[ghi]perylene

My settings are:
GC-MS: ThermoQuest GCQ Plus - ion trap quadrupole
Column name: Phenomenex Zebron ZB-5MS
Composition: 5 % Phenyl 95 % Dimethylpolysiloxane
Size: 30m x 0.25mm x 0.25µm
Ramp: 40 °C - 4 min, 240 °C - 25 °/min, 310 °C - 5 °/min, 310 °C - 10 min
Injector temp: 300 °C
Columns flow: He, 40 cm/sec
Injection type: Splitless
Needle rest in injector: 3s
Injection volume: 1µl (tried with 2 µl and 3 µl)
Transfer line temp: 300 °C
Ionization chamber temp: 240 °C
Standard: MRC with 1 mg/l (1000 ppb)

Tried both full scan and SIM, for SIM i used 4 time windows, with the ions for every PAH (Parent ion, Quantization ion, Qualifying ion)

Starting with Naphtalene and ending with Crysene everything is beautiful. Big peaks, good shape, good separation.

Benzofluoranthene elutes together with Benzo[k]fluoranthene in a very ugly shape, broadened. Same ugly, broadened and noisy form for Benzo[a]pyrene. But still identifiable and quantifiable.

However Indeno[1,2,3-cd]pyrene, Dibenzo[a,h]anthracene and Benzo[ghi]perylene are nowhere to be found. I tried A LOT of settings, with multiplicator, AGC, waveform filtering, scans per second. No luck.

I shortened the column, changed septa, liner, with new ones... nothing. Every other peak is very beautiful and strong S/N wise. Not the last PAHs...
Calibration is flawless, good signal, multiplier is auto set on 1300. Pressure in system is at 40mTorr, no leaks, no contamination.

Is it possible to be a column fault ? If not, what can it be ? Any advice would be appreciated.

Best regards,
Vlad Popovici

Try taking your final temperature all the way to 330C before your final hold, it may be that you are not getting hot enough to elute the compounds quickly enough. It will also help with peak shapes for the Benzo(a)pyrene if it elutes during the ramp and not the final hold.

[wraymogg]

Hi,

max temp for this column is 320. I forgot to mention I did that. I replaced 310 with 320 and the result was the same. I can try to raise the temperature from 240 to 320 faster, if you say it will improve the shape of the Benzo[a]pyrene.

Since the detector works very good for the other PAHs, i imagine is not a detector problem, but a setup or column problem.

Any tips to improve sensitivity ? In SIM mode i only got 1*10³ peaks. For 1µl of 1000 µg/l is little i think. Is like i loose analyte somehow, or the sensitivity of the device is really low ?

Thank you
Vlad

[James_Ball]

Hi,

max temp for this column is 320. I forgot to mention I did that. I replaced 310 with 320 and the result was the same. I can try to raise the temperature from 240 to 320 faster, if you say it will improve the shape of the Benzo[a]pyrene.

Since the detector works very good for the other PAHs, i imagine is not a detector problem, but a setup or column problem.

Any tips to improve sensitivity ? In SIM mode i only got 1*10³ peaks. For 1µl of 1000 µg/l is little i think. Is like i loose analyte somehow, or the sensitivity of the device is really low ?

Thank you
Vlad

We normally run our low calibration standard at 5mg/l concentration. 1mg/l will be difficult to see especially for the last compounds. One thing we did try and it seems to work is use a 5:1 split injection instead of splitless, it seems counter intuitive but it does raise the response of the analytes somewhat, especially the heavier compunds, and improves peak shapes.

[Bigbear]

I run about the same column ( Agilent) and the cpds. in question come out befort 311C.

[wraymogg]

Hi,

thank you for the split tip, I will try it tomorrow.

I also injected 10 mg/l standard, the result was almost the same. Actually you can see them at this concentration, as very very broadened and tiny slopes. I don't know why the peaks are so broadened and loosing shape in the end... On the column specification they are incredible good looking peaks. I tried A LOT of programs. On cold injections they will not elute AT ALL. On very fast ramps to 320 the noise and bleeding are very high.

What could be the solution to these very broadened peaks ? Should I try to increase the He velocity ? Now is at 40 cm/s.

Also i forgot to mention I am using Acetonitril as solvent. Can be a problem is not a non-polar solvent ? The peaks till Chrysene are perfect. After that the broadening begins and it gets to be for the last PAH almost one with the noise.

If I cannot detect even in SIM mode the last ones for 1 mg/l how can I get to the target of 2 ug/l ? Actually the law limit is for 0,002 ug/l but I can concentrate 1000:1. So i need to get the MS to 2 ppb (2 ug/l).

Best regards,
Vlad

[Steve Reimer]

Have you tried DCM instead of acetonitrile? With a 40 degree start temp and a boiling point of 81 you can have a liquid plug at the head of your column.

[01310040231]

I have not had good experience with Phenomenex. You may want to try the 8270D column from J&W using 0.18 diameter instead of 0.25. The compounds you are having problem with are heavier PAHs and you are not getting enough of it through the injector. The Agilent systems offers a pulsed injection meaning staring with high pressure to get more compounds into the column. This will make a significant difference for your sensitivity. Finally, try 4mm single goosneck liner with glasswool with 1ul injection. Good luck.

[James_Ball]

Hi,

thank you for the split tip, I will try it tomorrow.

I also injected 10 mg/l standard, the result was almost the same. Actually you can see them at this concentration, as very very broadened and tiny slopes. I don't know why the peaks are so broadened and loosing shape in the end... On the column specification they are incredible good looking peaks. I tried A LOT of programs. On cold injections they will not elute AT ALL. On very fast ramps to 320 the noise and bleeding are very high.

What could be the solution to these very broadened peaks ? Should I try to increase the He velocity ? Now is at 40 cm/s.

Also i forgot to mention I am using Acetonitril as solvent. Can be a problem is not a non-polar solvent ? The peaks till Chrysene are perfect. After that the broadening begins and it gets to be for the last PAH almost one with the noise.

If I cannot detect even in SIM mode the last ones for 1 mg/l how can I get to the target of 2 ug/l ? Actually the law limit is for 0,002 ug/l but I can concentrate 1000:1. So i need to get the MS to 2 ppb (2 ug/l).

Best regards,
Vlad

Most GC/MS analysis for these use DCM as the solvent, Acetonitrile is normally used when analyzing by HPLC/UV/FLD, which works great for that level of sensitivity. We normally can not see below about 5ppb for PAH compounds after 1000:1 concentration on GC/MS, but we use the same concentration step and have detection limits easily at your target level on HPLC. The levels you are attempting I see mostly only when using GC/MS/MS equipment.

[wraymogg]

Hi,

first of all, thank you all, a lot of what you say make sense and I actually discovered some flaws (40 degrees with acetonitril)...

I am using this solvent just because I bought the HPLC standard for PAHs. At that moment we weren't decided if to invest in a HPLC or MS for PAHs, so I bought this standard because it was suitable both for HPLC or MS.

Just today I received another transfer line, twice as short as the one I have now. It seems this shorter one is the original and proper one, so tomorrow I will install it. I don't know if this could be part of the problem since I tried the old one even at 320 degrees.

I don't have a PTV so I cannot inject more than 2 ul or I might flood the column. Also I don't know how to make a pulsed program, but I will try to see if I am able to vary the pressure. Also as recommended by 01310040231 I will buy a new 8270D column from J&W using 0.18 diameter. The column you have is 20m long ? My liner looks like this:



My column end is like in the picture. Is it set up correctly ? Or should I stick it into the neck ? Maybe is a better idea not to use wool in liner ?

My limits are the following:
benzo[a]pyrene - 0,05 µg/l
benz(b)fluoranthene + benzo(k)fluoranthene - 0,03 µg/l (sum)
benzo[ghi]perylene + indeno-(1,2,3-cd)-pyrene - 0,002 µg/l (sum)

My device is MS/MS capable (ion trap with MSn) but I never run a MS/MS analysis. Also I understand the sensitivity of a ion trap is way lower than a triple quadrupole. But I will try a SRM analysis.

I am highly handicapped by the software since I am using the original Finnigan software. I have both the Chemstation D.03.00.611 and Xcalibur 2.2 but I cannot import the old and weird Finnigan .MS file types. The only one accepting it is AMDIS.

On the 0.18 mm column I should never inject more than 1 µl, right ? Also you keep the injector at 300 degrees ? Usually the last one elutes around 26 min ?

Just to see exactly how my chromatogram looks, here it is: (taken in AMDIS since is the only one I can use to identify correctly)



Thank you,
Vlad

[James_Ball]

We are using the Restek Rxi-5SilMS column

http://www.restek.com/chromatogram/view/GC_EV01077

It gives us good separation on the Chrysene/Benzo(a)anthracene pair and we normally get about 30-60% valley on the Benzo b and k isomers.

With the tailing you have at the end of the run, maybe the column is too far into the inlet. Not sure on your instrument, but on Agilent instruments you only want the end of the column about 2mm into the inlet. One way you can check that is to leave the liner out of the inlet, insert a cotton swab that had the end flattened by tapping it on the bench, then watch for the stick of the swab to move when the column touches it. You can lightly tighten the nut until you can slide the column with some resistance, then move it up and down to get it about 2mm inside the inlet.

With that low response and severe tailing it seems the compounds are not being vaporized quickly enough for them to enter the column in a narrow band. Or maybe the Acetonitrile is causing it because it is spreading the band while the oven is cold.

[Peter Apps]

Try increasing the inlet temperature, and whatever splitless time you are using, make it longer; I suspect that the heaviest compounds are not leaving the inlet fast enough to get onto the column before they get flushed down the split line when the splitless time is up.

Peter

[wraymogg]

Hi,

I tried even 320 degrees on injector with the same result. Also is not because of Acetonitril since I started on 80 degrees too with the same result. However I did something weird. I didn't let the split valve open, and I didn't configured an open valve time since it was all the time manually closed. Basically I went all the time with pure splitless. This, I guess, forced all my solvent into the column since I didn't vented it at all.

It seems my problem has a name - Mass discrimination

Somehow I manage not to introduce enough of my high weight molecules.

I found this picture in a liner presentation, a liner that is very similar to mine. This is why I asked if i did wrong not to stick the column into the neck of the liner, because in the picture is stuck into the neck. Mine it is about 2mm into the inlet, but not stuck into the liner's neck. Here it is the picture:



I also don't have a Press-Tight seal like in the picture.

Best regards,
Vlad

[Peter Apps]

Hi Vlad

I would not expect any definite improvement with just a 20 C increase in inlet temperature - go to 350 C or even hotter.

The difference in the amount of solvent that gets onto the column with the split vent permanently closed compared to openening it after about a minute is negligible, and solvent effects would not cause the problem that you have.

Uniliners are designed to have the column sealed into the bottom taper, but with the split vent permanently closed I doubt that not having it sealed could cause the problems that you have. For the PAHs to elute as very broad peaks means that they are leaving the inlet as broad bands - anything that the column does to make those peaks wider would also affect the other PAHs whose peak shapes are good.

Peter

[CE Instruments]

You have a Thermo GC and Ion trap and yet the liner and injector diagrams all refer to Agilent injectors
Why do you have that liner ? When I worked for Thermo they had their own splitless liner.
The column should go 45mm into the injector which would put it just above the taper on the correct splitless liner. Your column is almost certainly too low in the injector.
Note the Thermo injector does not have a metal disk at the bottom and in correct usage will not suffer from discrimination.
You mention a 3 second dwell time (?) in the injector. Is this post or pre injection ? The Thermo injector works best
with a "hot empty needle " technique so a time of 2 seconds pre injection and 3 post is best.