Unable to get reproducibility in my peak areas

[jvtruebl]

Hi everyone, I've been working on this for about 3 or 4 months now and am my wits end. Any help is greatly appreciated. I'll try to include as much info as I can.

I'm trying to detect trace amounts of gaseous formaldehyde. To do this, I'm coating Tenax packed thermal desorption tubes with PFPH to derivatize the formaldehyde. Obviously before I start my experiment, I need to calibrate. To make calibration standards, I make a .0125% solution of 37% formalin in 20 ml of hexane. I then take a 1 liter tedlar bag filled with ultra pure nitrogen and inject various amounts (ranging from 1-3 ul) of my formaldehyde/hexane solution. The formaldehyde instantly volatilizes and I then connect my PFPH soaked desorption tube in-line with the tedlar bag, and sample all the air within it at a flow rate of about 50ml/min.

I then put the desorption tube in my desorber unit, and analyze via gc/ms. For whatever reason, I cannot get good precision from one run to the next. Below I've added a chromatograph of a few runs, with 2 nanomoles and 4 nanomoles shown.




Here are the specs of my method.


Method:

GC/MS: Varian 450-GC coupled with Varian 220-MS
Preconcentrator: CDS 8000 with vocarb trap. I'm using the tube desorber unit. Tubes are desorbed for 15 minutes at 300C. Then Trap then Desorbs at 275C for 10 minutes. The GC transfer line is set at 250C
MS METHOD: solvent delay til 10 minutes followed by scanning at 50-350 til the end of the GC run.
GC METHOD: 50C for 4 minutes, then 7C/min til 100C, followed by 8C/min til 250C. Hold at 250C for 2 minutes.
Flow rate is 8psi
Injector is at 270C with an initially splitless method until .75 seconds at which point it is split at a ratio of 1/20 until the end of the run.

I'm not sure if it's a GC problem or a preparation problem. Judging by the 'cleanness' of the peaks, I would say the GC/MS is doing it's job correctly. But that makes me wonder what is wrong with the prep? I checked, and there is no carryover between runs, the tubes are always thoroughly conditioned, the bags are always fully purged with nitrogen between runs, and I made sure all HCHO was being sampled out of the bags.

This leads me to believe it's the way that I put the HCHO sample into the bag. Is there a precise way of sampling with hamilton gas-tight syringes that I'm not following? Any clues would be appreciated. Let me know if you need more info.

[tom jupille]

Given the shift in retention time as the peaks get bigger, I'd guess that you're looking at an overload problem: i.e., it's not a GC problem but a mass-on-column problem. Beyond that, I'm out of my depth!

[Yama001]

I would be worried about the formaldeyhe absorbtion onto the TD bag. Perhaps the nitrogen purge should be added to the tube as well? Or is that too much gas volume - its been a while for me but I recall there are some breakthrough concerns.