TPH-Diesel Range Organics method questions

[pacerlaser]

Hello, I am trying to improve our current Diesel and Oil GC method. I have an Agilent 6890 GC with a ZB-5 30m 0.25 df and 0.32 ID using helium as the carrier gas. I would like to switch the carrier gas to hydrogen and improve the run time. Currently we have a 22.4 minute run that starts at 65 C and goes to 340 C. We use a retention time standard at the beginning of each run that contains many straight chain alkanes including decane, dodecane, C22, C24, C36 etc. The chromatography of these peaks is rather poor; I'm seeing a split peak for decane and peak fronting for the others. I use a 2 uL injection of diesel and oil in DCM as the solvent. We have a reporting limit of 50 mg/L on column for diesel and 100 mg/L on column for oil. I have tried using a ZB-50 column, 30 m 0.53 ID and 1.00 df and the main problem with this is that there is a baseline rise that happens when the column reaches about 260 C. I have to use an area sum integration in Chemstation so the baseline rise is very problematic particularly in the oil region of the chromatogram. Can anyone recommend a column that minimizes baseline rise and will prevent peak fronting for the heavier straight chain alkanes? Thanks for the help in advance!

[Peter Apps]

The peak shapes you describe sound as if you have solvent effects - which is odd given your programme starting temperature relative to the boiling point of DCM.

What is the concentration of the standard - and how much of each alkane are you putting onto the column ?

Is there a particular reason to use a 2 ul injection ?

What are your inlet conditions ?; temperature, split, liner type ?

Peter

[pacerlaser]

Thanks for the reply Peter, The inlet is at 300 C and it is a splitless injection. The standard is at 500 mg/L for each of the alkanes. If we cut down on our 2 uL injection volume we have difficulty reaching the 50 mg/L diesel reporting limit. The liner type is a 4 mm single taper with glass wool near the bottom. We use 5 mL/min constant flow through the column with a starting pressure of 16.9 psi.

[Peter Apps]

Thanks for the reply Peter, The inlet is at 300 C and it is a splitless injection. The standard is at 500 mg/L for each of the alkanes. If we cut down on our 2 uL injection volume we have difficulty reaching the 50 mg/L diesel reporting limit. The liner type is a 4 mm single taper with glass wool near the bottom. We use 5 mL/min constant flow through the column with a starting pressure of 16.9 psi.

You are putting 1000 ng of each alkane onto the column. That is a lot in each peak, and with multiple alkanes might be enough for them to interfere with each others peak shapes.

Try diluting the standard by twenty to get 50 ng per peak.

To speed up the analysis you could also look at using a shorter column rather than running the carrier gas as such high flow rates.

Peter

[Hornet]

Hello, I am trying to improve our current Diesel and Oil GC method. I have an Agilent 6890 GC with a ZB-5 30m 0.25 df and 0.32 ID using helium as the carrier gas. I would like to switch the carrier gas to hydrogen and improve the run time. Currently we have a 22.4 minute run that starts at 65 C and goes to 340 C. We use a retention time standard at the beginning of each run that contains many straight chain alkanes including decane, dodecane, C22, C24, C36 etc. The chromatography of these peaks is rather poor; I'm seeing a split peak for decane and peak fronting for the others. I use a 2 uL injection of diesel and oil in DCM as the solvent. We have a reporting limit of 50 mg/L on column for diesel and 100 mg/L on column for oil. I have tried using a ZB-50 column, 30 m 0.53 ID and 1.00 df and the main problem with this is that there is a baseline rise that happens when the column reaches about 260 C. I have to use an area sum integration in Chemstation so the baseline rise is very problematic particularly in the oil region of the chromatogram. Can anyone recommend a column that minimizes baseline rise and will prevent peak fronting for the heavier straight chain alkanes? Thanks for the help in advance!

Hello,
i read many thing i don't like in your method:

- the first thing i would change is the solvent. DCM has as a big vapor volume, i use heptane for the same method but hexane would work as well, try injecting the same volume in the same conditions just with hexane or heptane and see what happens. 65°C is good with hexane too, raise the initial temp to 80 °C for heptane.

- 2 ul splitless and your liner are good and that's probably the best injection type you can do with a split/splitless injector, but what is the splitless time? I tried many solutions years ago, i ended with a splitless time of 1 min which gave a good solvent shape and low detection limit (~15 mg/L of 1:1 diesel/oil mix)

- Now the column...you column is ok, i did this analysis for 3 years with a ZB-5 30 m (mine was HT inferno) until i discovered how much time i was wasting with a long column.
Anyway, my program with a 30m ZB-5 was:

2 ul splitless single gooseneck with wool liner, 1 min splitless time.
280°C injector temp. (300 is fine too)
oven: 80°C (heptane only) for 1 min, 20 °C/min to 340 °C, 340°C for 16 mins - Total time 30 min.
FID temp: 340 °C
Helium flow: 25 psi constant pressure (which is around 2 ml/min at 80°C), 5 ml/min are you sure? that seems a lot.

- last point is...why do you care about how much the baseline rises? The only important thing is that it is stable once it gets to 340 to have a flat baseline for all the final temperature time. Everyday you start your instrument you have to do a blank run first, to clean the column, and your second run is your baseline that you have to subtract to every sample you'll inject after.
If you have a stable baseline run once you subtract it your sample will be a flat line with only analyte peaks...Your column is good, forget ZB-50 which is used for other types of analysis.

- once you take your 30m ZB5 to the end, look for a 15m ZB5-HT, i'm using it from 1 year and my runs are 10 min faster.


Other than that remember that in this analysis the less you separe your peaks the lower your detection limit is. Try to make all the peaks to be in a narrow amount of time and you'll see an higher "curve" to integrate.

[AICMM]

pacerlaser,

I vote with Peter, you are overloading your column. You could switch to a different solvent but 2 uL of DCM is not too much for a 4 mm id liner on expansion volume. But 2 uL to hit detection limit raises the big questions and 1000 ng on a 0.32 column is too much.

Have you considered column compensation?

Best regards,

AICMM

[twranger]

This should be a very simple method, but can't quite pinpoint your problem. Some possiblities:
we use single taper liners with wool at top, 1ul injection
there should be little to no baseline rise unless there is a leak or column phase shot
what's your oven ramp rate? we are at 12C/min, with run time of 34 minutes. Hard to avoid with the heavies in this method.

We use the DB-5 column, helium carrier at 15 psi, nitrogen make-up to FID. I recently started using a 5:1 split because I was losing the later alkanes, but peaks look much better.
Diesel standards are 100-10000 ug/ml, which is fine for 1liter/20gram samples concentrated to 1ml.

Good luck, TW

[antonk]

As far I know Diesel have to be analysed on non-polar phase

Helium carrier 100m
http://www.unichrom.com/lib/gce/a27.pdf

Hydrogen carrier
http://www.unichrom.com/lib/gce/a28.pdf

[pacerlaser]

Thanks for all the replies, I'm finally getting around to taking another shot at this. Its been a busy summer. I'll let you know how it goes, thanks again!