[GCxGC-MS] 2D chromatogram curves in one sample but not the others

What would cause this kind of curving on the chromatogram?

The samples below are from olefin oligomerization experiments done with different reaction temperatures, so they have different compound compositions and aromatic contents. But shouldn't these images be mostly the same shape when the same method was used and the samples don't differ that much from each other? The curved sample was run again to see if the curving repeats, and there was no difference.

Sample 1.

Sample 2.

Sample 3. Suddenly there is a lot of curving and wrap around.

The method that was used in all three samples is presented in the table below:

GC conditions
Column set First dimension: DB-5MS Ultra Inert (20m, 0.18mm, 0.36µm)
Second dimension: DB-HeavyWAX (4m, 0.18mm, 0.18µm)
Carrier Helium, constant flow
First column: 0.5 mL/min
Second column: 7.5 mL/min
Inlet 0.1 µL, 250C, split mode, split ratio 1:250
Oven 40C (5), 3C/min, 280C (5)
MS Transfer line 280C
Modulator conditions
Modulation period 2.50s
Inject time 0.25s
MSD conditions
Acquisition type Scan
Mass range 35-500 m/z
Scan speed 20,000 u/s

The GC used in these analyses is Agilent 8890 with Agilent 5977B MSD. The modulator is Agilent G3673A Reverse Fill/Flush CFT Differential Flow Modulator.

  • Hi kalle-h, 

    Has the method worked well for you in the past?  Any changes recently like column trims?  Sometimes small pieces of column can get caught inside the modulator fittings and impede flow, which has negative effects on our data.  

    If your peaks are consistent along the X-axis, there probably isn't an issue in the 1st D.  This looks like a modulator or 2nd D column change.  The compounds are eluting more slowly off column 2, which is causing them to display higher in the Y-axis when the data is processed.  The intensity of peaks looks lower in sample 3 compared to sample 2, so I would say we are losing sample at the modulator so it isn't registered at the proper time on the detector.  It seems to affect the entire run, not lighter or heavier compounds specifically.  

    This is the troubleshooting process I would take to figure out where the problem is:  

    If you run a solvent blank, do you see peaks?  There could be carry-over in the system.  When you do a library search against the peaks that are showing up in the expected blank space of images 2 and 3, what compounds are identified?  Are they expected as part of your sample or does it look like contamination? I could speculate that the extra peaks that are showing up are the missing parts of the lower intensity compounds, but I can't say for certain.   

    How consistent is your unprocessed data - not the GC Image processed version, but what comes from your data system?  Could this be a setting on GC Image that is rendering differently?  

    Your modulator settings look good, we just need to make sure that the system can supply the flows the method is calling for. 

    Do you have a post 2D column split between another detector and your MSD (commonly an FID)?  What does that detector chromatogram look like? Ideally you would see the exact same profile if the splitter is configured properly, but it may give us a clue.  If you have a second detector, I would recommend measuring the column flows with a flowmeter to verify the configuration is delivering the the same values as the GC setpoints.     

    There are many other things to investigate, but this is where I would start.  

    Let us know what you find out?


Was this helpful?