Linearity/Recovery

[youngster_mystra]

I am trying to validate a method, for determination of methanol by gc in an active pharmaceutical ingredient. I am using ethanol as an ISTD (300 ppm).
When i check linearity with spiked standards everything works fine.
When i am spiking real samples, the recovery that i got in 300 ppm of methanol was not good. The runs were made in four different samples and used the same chromatographic conditions as in the linearity runs.
For some reason when methanol/ethanol ratio is 1/1 in the real sample, the recovery is not good.

Any comments or ideas?

Thanx in advance

[Peter Apps]

Do you see the same poor recoveries at other concentrations of ethanol and methanol in real samples ?

Peter

[youngster_mystra]

No, until 200 ppm of methanol it's all ok.
And from 400 to 500 ppm of methanol works fine.
There seems to be a gap in the recovery from 200 to 400 ppm.
The choise of the concetration of the ISTD was made randomly and i am wondering if there is a problem between alcohols when the ratio is 1 to 1.

[Peter Apps]

At ppm concentrations I doubt that there is an interaction between the alcohols. Having the effect only at the 200 ppm level makes it likely that there was an error in diluting a standard, or spiking a sample.

Peter