water problem with USP 467 residual solvents

[liyu699]

HI everyone,
Our company just equipped a PerkinElmer GC together with H40 Head-space for USP 467 residual solvents analysis.

Now I am able to apply the USP method only when water is not added to headspace vial (according to USP 5ml of water is supposed to be used)
When water is not added (only residual solvent standards in DMSO) all peaks come out perfectly separated and peak shape everything is great.
However, if add the 1ml sample/standard into 5ml of water for analysis, peaks are co-eluting and peak shapes are awlfull, especially at the beginning half of the chromatogram.

I tried increase the needle temperature and transfer line temperature to 110C and 125C respectively, to eliminate the possibility of water condensation, however, the result is the same.

Dose anyone have the same experience?

[alphananotech]

What parameters of the FID are you using?

[cromatoloco]

I have problems with the same method, I use the USP std mixture, class1, class 2A and class 2B in water and DMF (separately). In the class1 my problem is that I obtain a signal to noise for CCl4 = 2.7 (NLT 3 according USP) and in th e class 2A the R=1.02 (between acetonitrile (ACN) and methylene chloride) (NLT 1.0) so I am in the limits and I did many changes to obtain the best results.

Conditions:
oven: initial temperature 40 for 20 min rate 10C/min final temperature 240 for 20 min. Column 30m x 0,32 mm x 1,8 um. Constant flow = 2.2 mL/min. split ratio=8. Liner 4mm split. split septum purge = 3. makeup gas flow 20 mL/min He. Carrier gas He. Inlet 140C. Detector 250 C. Headspace vial pressure 10 psi. Vial pressurization time 1 min. Equilibration time 45 min. Oven temperature 85 C. Transfer line 120C. Loop 105C. Cycle time 65 min. Headspace vials 20 mL and 10mL no difference in vial size.

To lower split ratio's more signal to noise (S/N) and lower R. For example to split ratio=5, S/N=8 and R=0.3 and split ratio = 10, S/N=2.8 and R=1.2.