Chromatography Forum

I'm attempting to run a residual solvent method that was developed for GC/FID on an HP6890 w/ HP5973 MSD.

The method in question has been run on a 6890 with FID several times before. The analytes are methanol and chloroform. On the FID instrument, methanol typically elutes @ ~2.2 minutes and chloroform @ ~4.2 min.

When I run the same method on the MSD, methanol is eluting @ 1.1 min and chloroform @ 2.2.

I've confirmed that the oven temperature program is the same, and the oven temp is accurate throughout the run. I've also tried more than one column. Retention times remained the same.

The instrument parameters are as follows -

Oven program:

T °C / Hold, min / Ramp, °C/min

50 / 3 / 20
150 / 0 / 30
300 / 10 / NA


Flow: Constant, 2.0 mL/min

Injector:

Temp = 200 °C
Carrier: Helium, split ratio = 20:1
Injection volume: 2 uL

Aux temp: 280 °C (Original method has FID temp set at 300 °C)

Column is a Restek Rtx-5, 30 m, 0.32 mm i.d., 1um film

Run Time: 23 minutes

The sample solvent is DMF. I know this isn't the best solvent to use for MS applications, but my hands are tied at this point. That said, the DMF peak elutes in just under 5 minutes on the MSD but is not typically observed when using FID because the method states that data should only be collected for 6 minutes. . . Another clue that everything is eluting too quickly on the MSD.

My question is this: Has anyone observed such a drastic decrease in retention when switching from FID to MS before? Is this behavior inherent to the technique (due to vacuum pulling at the detector end), or is some other explanation more feasible?

Thanks for any feedback.

-Joe

The one big difference is the vacuum at the outlet end of the column in the MSD. This will of course change the flow through the column and shorten your retention times. The Agilent GC has an option when you configure the column to turn on compensation for the vacuum which will partially help with the r.t. shift.

Regards,
Mark

The vacuum compensation for the MS should be on by default. The problem is the linear velocity is different. Check what the linear velocity of the carrier gas is on the GC FID system and set the GCMS to the same linear velocity and the retention times should be very close.

Just did a quick calculation, the required head pressure to set the linear velocity to the value to give the same retention times is a negative number. If you go to an open split interface you can probably get the same retention times, otherwise you would have to change the method to use a norrower bore column.

Just did a quick calculation, the required head pressure to set the linear velocity to the value to give the same retention times is a negative number. If you go to an open split interface you can probably get the same retention times, otherwise you would have to change the method to use a norrower bore column.

Thanks for the info. For my purposes, it isn't critical that the retention times be identical for both methods. I've got adequate resolution as-is, and as long as the retention time shift is more or less constant for all peaks, the difference shouldn't present much of a problem.

I appreciate everyone's help, nonetheless!