GC Headspace - measurment of ethanol

[BB63]

With a GC headspace to analyse ethanol in aqueous solutions from traces to % , is it better to use an internal calibration or external calibration technique.

As far as I know with a classical injector internal calibration is required...

Any ?

[Peter Apps]

Better in terms of what ? - accuracy, precision, cost, throughput, robustness etc etc ?

Peter

[BB63]

better in terms of robustness of course... this also means accuracy and precision doesn't it ?

My question is since internal calibration is the "state of the art" method when using a classical injector, do I have to consider the same for the same reasons with Headspace injection.

In other words, would HS injectors guarantee a perfect and representative quantification of the volume / analytes injected ?

Thanks for your prompt reply.

[AICMM]

BB63,

Is your matrix perfect? The same every time, no changes? If not, I would recommend an internal standard.

Best regards.

[chromatographer1]

Internal stds are not needed in my experience with headspace analysis.

Accurate and reproducible results are dependent upon the hardware involved and the care of the chemist in preparing samples and external standards.

Given the diversity of choices for an internal standard it is possible to make errors in measurement if the above requirements are not observed due to the differences in volatility of analyte and internal standard.

I have found with the PE and with the Varian (Tekmar) systems and I suspect with other major vendors that the headspace sampler to be more accurate than the operator preparing the samples, unless extraordinary measures are taken.

External standards MUST match the matrix being tested but without containing the analyte, a situation that is not always attainable.

It is suggested as a better procedure (in that it confirms the proper operation of the analyzer and the correct preparation of the samples) that a calibration based on a standard addition method is the preferred choice.

Errors may not be caught without this internal confirmation and it gives examiners confidence in the research if such confirmation is demonstrated.

You may email me with questions if I have been too terse or unclear.

Rodney George
consultant

[BB63]

Thanks Rodney for your advised answer.

quote "External standards MUST match the matrix being tested but without containing the analyte, a situation that is not always attainable. "

Do you mean that when measuring ethanol in a liquid extract containing sugars, polyphenols and some minerals the extarnal standard should be prepared without ethanol ?

I'm off for a long weekend vacation....

'Be back soon.

[Peter Apps]

No, robustness does not also mean accuracy and precision, it means the ability of the method to return the same measurement result (within specified limits) despite small (and sensible and deliberate) changes in the operating conditions.

Headspace analysis depends on partition of the analyte between the sample and the headspace. An ideal internal standard would have to partition in the same way as the analyte, in other words it would have to have the same molecualr weight and the same functional groups. Unless you can work with optical isomers this is not chemically possible.

You CAN use an internal standard to correct for differences in non-partitioning aspects of the method such as injection volume and split ratio.

Peter

[chromatographer1]

Good answer and well stated, Peter.

Have a nice long weekend, BB63

Rod

[krickos]

Hi

Agree with Peter here.

If you want to be sure to exclude any impact from matrix effects (ie sample) use standard addition instead of external calibration. Matrix effects from samples in water solutions are quite common and can be quite significant (spiked samples with salts in it can give> 10% higher respons than equvivalent external standard)=> know as the "salting out" effect

[Peter Apps]

Hi Rod, you are kind of quiet these days

"Accurate and reproducible results are dependent upon the hardware involved and the care of the chemist in preparing samples and external standards.

Given the diversity of choices for an internal standard it is possible to make errors in measurement if the above requirements are not observed due to the differences in volatility of analyte and internal standard.

I have found with the PE and with the Varian (Tekmar) systems and I suspect with other major vendors that the headspace sampler to be more accurate than the operator preparing the samples, unless extraordinary measures are taken. "

At my previous job, where we took extraordinary measures over everything because it was a national metrology lab, I worked up a very repeatable headspace ethanol method by drastically modifying an Agilent G1888. I got down to about 0.3% rsd on raw peak areas, and a bit below 0.2% on ethanol area compared to butanol as an internal standard. This was under the strictest possible repeatability conditions; vials charged with portions of a bulk solution containing both ethanol and butanol, not made up individually. My colleagues carried on after I left and found that in the real world scenario of making up vials with separate samples, each spiked with internal standard, the rsd of standard addition was indeed higher than the instrument repeatability, just as you say, even with extraordinary measures. I'm not sure how far they got with it - I must e-mail them and see how it went.

I should maybe add that out of the box neither the G1888 or a Gerstel combi pal clone was more repeatable than the standard addition step under the extraordinary conditions of a metrology lab (we used gravimetric checks nearly every time we added anything to anything, and corrected all results for actual masses used). It was fun in a strange kind of way

Peter

[rbardsley]

Are you performing static or dynamic headspace? Dynamic would be better for trace and static would be better for %.

[Ron]

One thing I have noticed is that internal standard addtion for headspace helps compensate for variations in vial sealing as long as an appropriate internal standards is used. There are large variations in vial sealing from operator to operator in many cases using manual crimpers, so I am a fan of the electric crimpers. In many labs doing headspace the limiting factor in precision using external standard is vial sealing.

[krickos]

One thing I have noticed is that internal standard addtion for headspace helps compensate for variations in vial sealing as long as an appropriate internal standards is used. There are large variations in vial sealing from operator to operator in many cases using manual crimpers, so I am a fan of the electric crimpers. In many labs doing headspace the limiting factor in precision using external standard is vial sealing.

Good point, "basic" thing that sometimes is overlooked. There are also "pressured air" operated crimpers with foot pedals, like those even better than the electic ones, even though not porteble great when you have a large sample flow (at least the early electric porteble ones was a bit heavy when doing many samples).