Method to separate C7-C40 alkane ladder/peak tailing problem

[u0008355]

Dear all,
I have a Thermo ISQ GC/MS with a Trace GC 1300 GC, a Restek MXT-5 (ie DB-5 HT) 30m 0.25 mm ID 0.25 µm film thickness column and use Agilent 4mm ID tap GW (part nr 5062-3587) liners and He as a carrier gas, and use liquid injection (hexane solvent, injection volume 1 2 µl). I am on the lookout for a good method file to separate a C7 to C40 alkane ladder, resulting in sharp, approx equally spaced peaks, no peak tailing and with also the longer chain alkanes coming out nicely. [Ideally I would like to get compounds out in the range of C7 to C45, but my alkane ladder only goes up to C40]

The method and settings I am using now (thermo Trace GC 1300 methods file available from here: https://www.dropbox.com/s/mohx1685d4gb2 ... .meth?dl=1) are pasted below, but as one can see in the chromatogram my alkanes with chain lengths above C35 show peak tailing :



Anybody any thoughts how I could improve resolution for these long-chain compounds, and get rid of the peak tailing there? Any obvious improvements I could make to my method?

Current settings for my method are:
- Restek MXT-5 (ie DB-5 HT) 30m 0.25 mm ID 0.25 µm film thickness column
- Agilent 4mm ID tap GW liner
- liquid spitless injection with surge (without surge the longer-chain alkanes dropped off too much)
- 1 µl injection volume
- constant flow, 0.9 ml/min
- 350 °C inlet temp
split flow 80 ml/min
splitless time 0.8 min
- surge pressure 170 kPa
surge duration 0.8 min
- purge flow 5 ml/min
constant septum purge
- carrier options
vacuum compensation on
carrier gas saver on
gas saver flow 20 ml/min
gas saver time 2 min
- oven programme
40 °C for 2 min,
40 to 120°C at 20°C/min
120 to 200°C at 10°C/min
200 to 250°C at 7°C/min
250 to 350°C at 5°C/min, 4 min hold at 350 °C
- MS transfer line temp 350 °C, ion source temp 300 °C
- injection settings:
35 mm injection depth, 20 mm/s penetration speed, 50 µL/s injection speed
fast injection unchecked

Any thoughts what could cause this peak tailing for the long-chain compounds? Or could that be more of a liner issue?

cheers,
Tom

[MMJ88]

With low volatility compounds and splitless injection, you aren't generally going to get a quick transfer of compounds from the liner to the head of the column.

One thing you can try is inject a standard repeatedly, cutting down the splitless time slightly each time. You may find a compromise between sensitivity and peak shape.

[u0008355]

Thanks for the advice - I was intruiged though by the fact that my short-chain peaks were all well focused, and that the peak tailing only affected the compounds with chain length >C35. Any thoughts what could be the reason of that? Still doing some further tests with shorter splitless times & surge durations (I always take the two the same, is that what I should do? - in a first try a surge duration & splitless time of 0.3 mins at least gave sharper peaks, but still gives peak tailing for the long-chain compounds) and higher surge pressures. The pulsed injection seemed to help a lot to get the peaks sharp, and have everything transferred onto the column as quickly as possible... I also tried split injection with a 1:10 or 1:100 split, and that also was not significantly better, and still shows the peak tailing for compounds with chain length >C35.

[Yama001]

We routinely do that separation (C8 to C40, including the odd alkanes, pristane and phytane as well) on an FID with 30m by 0.32 mm ID DB-5 column.
Start at 60 degrees, hold 1 min, ramp to 320 at 25 deg/min, hold until 12 min.
Carrier is hydrogen at 6.2 cc/min, 5:1 split, inlet at 300 degrees, 4 mm ID liner - gooseneck at the bottom, glass wool around two thirds down.

I have in the past run a dual MS/FID istrument, with the same column on the FID, but using the 30m by 0.25 mm ID column for the MS.
This is all Agilent 6890/FID and 7890/5975/FID.

It was pretty annoying to see the GC/MS run look poor compared to the FID run. It appears the vacuum outlet changes the chromatography
significantly. It also looks like the 0.32mm ID DB-5 (likely DB-1 works as well or better) appears to work best for this separation.
I think some folks have mentioned putting a narrow bore restrictor (0.1mm ID?) on the end of the MS column to reduce the effects of the vacuum sucking
at the end of the column.

[Rndirk]

Is your analytical column inserted into the MS or is there an extra piece of uncoated column?

Either way, having that piece at a temperature of 350°C all the time could damage it and might be causing problems with tailing of heavier compounds.

300°C should be OK, condensation of heavy components will not happen there, the vacuum of the MS increases the velocity in this part of your system.

Another idea is to increase the constant flow of the method, if you are allowed to change the retention times.

[u0008355]

Thank you all for the replies! Indeed going from a GC to a GC/MS method one has to recalculate flow rates etc to take into account the vacuum, but the separation power / resolution should be the same I believe... I also tried a flow rate double as fast with ramps twice as fast, but resolution did not get better (guess I would have to use a shorter column and thinner film then). So it's not a diffusion problem with the long-chain compounds. Increasing the surge pressure and shortening the surge duration and splitless duration a bit seemed to slightly help to get more narrow peaks, but not for the long-chain compounds. I start to think it could be a liner problem - right now I am using a straight-through 4 mm ID one - maybe I should use a 2 or 3 mm one, and maybe a single tapered one, maybe with a focusliner or a carbofrit?
If I use the SGE liner selection tool at
http://www.sge.com/products/gc--lc-supp ... et-liners3
and select Thermo, wide boiling point range, splitless injection, I end up with
092003 4 mm ID, single tapered with focusliner
092111 2.3 mm ID, single tapered with focusliner
For high-temperature programmes and splitless injection I also saw this one recommended:
http://www.restek.com/catalog/view/9370

Anybody any thoughts which of these could be best for this application and injection method?

[Hornet]

I don't think it's a liner problem...can i ask you why gc/ms?

[u0008355]

Well because my samples contain complex mixes of hydrocarbons (methyl-branched, alkanes, alkenes, alkadienes, etc) plus some other classes of compounds, which can only be identified and quantified based on a combination of retention indices and mass spectra!

[Yama001]

I have worked with people who insist that a straight 2 mm ID liner is best for this work, particularly to address apparent discrimination across the mas range (I think Gerstel takes this approach regarding splitless injections). I have seen good results using either (4 mm vs 2 mm, split or splitless). I would agree that using a pressure pulse (I assume this is the same as a surge?) can help a splitless injection.

[u0008355]

Thanks for the info! In my case splitless injection with a pressure pulse (is indeed what thermo refers to as a pressure surge) in combination with a low initial GC oven temp of 40°C was quite successful in eliminating discrimination against long-chain compounds (it's very close to being equal in fact). By contrast, split injection always showed strong discrimination against long-chain compounds. Together with the fact that I usually have fairly small concentrations I would therefore go with splitless injection. For the liners maybe I'll give them all a try and just see what I get... Hard to predict the result I guess...

[MMJ88]

For the liners maybe I'll give them all a try and just see what I get... Hard to predict the result I guess...

I think that's a good strategy - and switching liners is pretty easy. Hope you get it figured out. I would try a liner with glass wool and a taper at the bottom, I've used this one for my splitless methods - http://www.restek.com/catalog/view/11042. Try a couple types and see what works for you.

[environeer]

This peak tailing problem describes pretty well my current issue. How unfortunate that now further results got posted here...

[Consumer Products Guy]

See chromatogram below of paraffins; is this what you're attempting?