Wavy GC baseline on some samples


We are working on developing a method for detecting mineral oil contamination (MOSH and MOAH) in edible oils. We are able to get MOSH (mineral oil saturated hydrocarbons) to work well, but when we inject a sample for MOAH (mineral oil aromatic hydrocarbons we are getting a wave in our baseline during the ramp. I believe it bay have to do with the sample itself, because I don't see this on the blank injection or the MOSH injection using the same GC parameters. Has anyone seen this before or have any suggestions. 

The samples are in iso-octane. But the MOAH sample may contain traces of other solvents such as DCM and toluene. 

The GC is a 7890A with a PTV inlet configured to inject as if it was on-column using a Restek liner, Catalog No. 24977. Detector is a standard FID.

We are using a 15 m, 0.32 mm, 0.10 µm DB1-HT column with a 5m, 0.53mm retention gap/ guard column. 

GC parameters are as follows:

Equilibration Time 1 min
Max Temperature 400 °C
(Initial) 40 °C
Hold Time 2 min
#1 Rate 5 °C/min
#1 Value 50 °C
#1 Hold Time 0 min
#2 Rate 15 °C/min
#2 Value 350 °C
#2 Hold Time 8 min

PTV Inlet H2
Mode: Splitless
Pressure: On: 9 psi
Total Flow: On: 16.686 mL/min
Septum Purge Flow: On: 3 mL/min
Gas Saver: Off
Purge Flow to Split Vent: 2 mL/min at 1.5 min
(Initial) 40 °C
Hold Time 2 min
#1 Rate 5 °C/min
#1 Value 50 °C
#1 Hold Time 0 min
#2 Rate 15 °C/min
#2 Value 350 °C
#2 Hold Time 8 min

Detector FID
Makeup: N2
Heater: On: 370 °C
H2 Flow: On: 40 mL/min
Air Flow: On: 400 mL/min
Makeup Flow (Combined): On: 25 mL/min
Carrier Gas Flow Correction Included in Makeup Flow
Flame: On

  • I'm curious if this pattern changes with the type of oil analyzed?  It looks like a triglyceride pattern to me, which would make sense in an oil.  However, I would expect you to see the same pattern on your MOSH injection, provided the oil matrix is the same?    

    Is this a prepared sample in oil that has a CofA?  I would be interested to compare the properties/ranges of saturates vs aromatics.   Maybe there are some tweaks to give more resolution of aromatics from your matrix.  You could employ a phased pre-column, for example.  Something like a DB5 could hold onto the aromatics a little longer than the triglycerides.  It could lengthen your run, so you would need to evaluate the options.  

    I would imagine there are some cleanup steps out there that could make the analysis a little cleaner, if you have the freedom to pursue that route.  Otherwise, I would make sure everything is coming out in a single run, and there is no carry-over.  

    Great question and thanks for providing all the detail!


  • Hi Abbey

    Thanks for the info!

    So far we have only been running this on various forms of olive oil. But the waves in the baseline remain regardless of MOAH contant. I have run samples that were analyzed by outside labs (with COA) that contain 0ppm and 50ppm. The waves are identical. 

    Our MOSH results look fine and are giving us accurate results. 

    I like the idea of the db5 pre-column! I may try that, I would just need to make sure it wouldn't have any unwanted effects on MOSH because we run them on the same column. 

    I have been looking into the cleanup options. Currently our sample prep runs the sample through a silica/sodium sulfate/10%Ag silica column. The MOSH and MOAH are eluted with different solvents. MOSH with hexane, and MOAH with hexane:DCM:Toluene (75:20:5)

Was this helpful?