DMF peak in USP residual solvents

[simonhall]

Hi All,

We're analyzing pharmaceutical APIs using the USP residual solvents method. We're seeing a DMF peak in the sample but a much smaller peak (more a lump) in the spiked sample. This seems somewhat counterintuitive. The spiked sample is prepared the same as the sample but with 10ul of DMF standard added.

DMSO has been used in making up of all solutions so there's no introduction of DMF from other sources. Other components (Toluene, Methanol, Ethanol) have all had larger responses in the spiked sample.

The following parameter, all from the USP, were used:
Headspace Parameters 3:
Equilibration temperature: 80
Equilibration time: 45
Syringe Temp: 105

GC Conditions
Column 0.53mm x 30m x 3.0um (G43)
Flow 35
Split 1:10
Inlet Temperature 140
Detector Temperature 250
Initial Oven Temp 40
Initial Hold 20
Ramp 10
Final Temp 240
Final Hold 20

Any thoughts as to why we're getting variable peak response would be greatly appreciated.

Kind Regards,
Simon

[dlbenach]

Hello,

Sounds like you are running USP<467> Procedure C, Water Insoluble with Procedure A GC conditions. Please confirm. Also, can you give the details of how you are preparing your Standard Solution, Test Solution, and Test Spike Solution.

Dave

[simonhall]

Hi Dave,

We are indeed running USP<467> Procedure C, Water Insoluble with Procedure A GC conditions. Solution preparation is as follows:

Stock Test Solution: 500mg sample > 10ml DMSO

Working Test solution: 1ml stock solution + 5ml water

Working Standard Solution: 1ml DMSO + 5ml water + 10ul USP reference standard

Spiked Test solution: 1ml stock solution + 5ml water + 10ul USP reference standard

Since performing the analysis we've realized the standard solutions are too concentrated so in future will be putting in a dilution step though I don't see how this would affect the variable response.

Kind Regards,
Simon

[krickos]

Syringe Temp: 105°C

What is the boiling point of DMF and DMSO?

Easly forgotten that temperatures in the sample gas pathway should be above the bp of all solvents including sample solvent/solvents. Not always seen directly but residues of DMSO/DMF/DMAA etc tends to build up in syringe or valve based systems, causing sensitivity to drop, poorer peakshape etcetera.

The recommendations in USP/PH Eur seems to be based on flow based injection systems, less sensitive in that regard than valve/syringe based.

Finally, DMSO mixed with water tends to be less predicteble than Water/DMF. In fact what I have seen you will actually loose sensitivity for DMF in DMSO/Water mixes as water % increases.
Example:
http://www2.shimadzu.com/applications/gc/g228.pdf

[dlbenach]

How does the chromatogram of the Standard Working Solution (SWS) look? If that looks similar to your Spiked Test Solution then there is probably a system issue. If the SWS looks good, then maybe there are sample matrix effects going on. Your concentrations look OK by my calc. You state you are using Headspace Parameters 3. If so, then your syringe temp should be 80-90C. I doubt that is the problem though.

Dave