Chromatography Forum

Can anybody suggest calibration parameters for headspace GC instrument

They depend entirely on your chromatographic conditions and the resulting chromatogram.

Calibration parameters

You should reproduce peaks of a volatile impurity from a standard preparation or a duplicated sample and achieve a relative standard deviation of less than 5%, preferably less than 2%.

You should demonstrate that the recovery of a volatile from a sample and a sample is linear over a defined range of sample size and volatile contamination content.

You should be able to demonstrate that your sample preparation is highly accurate and determine the size of the resulting error created by known errors in sample preparation.

You should demonstrate that your results are reproducible over the time involved from initial sample preparation to the time of the analysis of the last sample or standard. In other words duplicated samples prepared at the same time but analyzed at different times (2 or 12 or 24 hours later) should give the same result.

If acceptable parameters like these cannot be met you will not be able to perform quantitative analysis of volatiles.

best wishes,

Rodney George
consultant USA

Hi

Not disagreeeing with Ron or Rodney.

In short other considerations

First try external standards, recovery will early on tell you if you have an matrix effect (common in headspace) and if recovery is accepteble or not.
If matrix effect is unacceptable one might work around it for instance adding 0,5 sodiumsulfate (especially if water is main diluent) to all vials and recheck recovery.

If matrix effect can not be worked around in an accepteble way or if your matrix changes from analysis to analysis shift to standard addition (usually more injections than external standards)

Even MHE (multiple headspace extraction) can be used to work around matrix effects, tend to require even more injections and personally not in favour of it for routine analysis.

krickos has made valid points.

I recommend since it eliminates a host of issues that std addition is the preferred method of analysis.

Once validation of the analysis is determined (from the list of concerns I gave previously) then only three preparations are required to be run in measuring volatiles in a sample. The sample itself and two standard additions of a range of concentration previously determined to be linear.

This will give you three points for a linear regression and if the regression is acceptable, ie. >0.995 to 0.999, you will have a proven x-axis intercept for your value of the volatile content.

Regulators will have no problems in accepting values from a valid three point determination of volatiles. What complaints of poor science can be made? Since a HS run should not require more than 20-30 minutes of total run time, a ssample should take about an hour to process, including all sample preparation time and calculations of the three sample results, certainly less than two hours for the worst case.

best wishes,

Rodney George
consultant USA

Hi

Thanks to all

It is planned to take two external standards and will be given continuous injections in different concentrations to check the correlation coefficient.

But is it correct to fix a sample matrix for the calibration purpose

Regards

Mageshbalakrishnan

Any calibration preparation should have the same identical matrix (less the analytes in question) as any sample.

For example, if you were testing aspirin tablets for acetic acid content the calibration mix would contain the same amount of water, and the same aspirin tablet formula (without any acetic acid present) and a known amount of acetic acid added to it.

Rodney George
consultant USA

Go for MHE, there is matrix not really critical. That is one of the advantages with Multiple Head space Extractions that there is no or very small matrix effects. We uses that for routine analyses.