I am trying to develop a method for the analysis of flavors using GC-MS. We have an HP HS sampler and my main question is how do I determine the equilibrium time needed to reach equilibrium???
Please help as soon as possible and give as much detail as possible.
By Anonymous on Tuesday, August 26, 2003 - 08:43 am:
I do it by running a series of experiments just increasing the equil time until there is no significant peak changes from the previous injection.
By Matthew on Tuesday, August 26, 2003 - 09:42 am:
Thank you very much Mark. That's kind of what I figured, but I just wanted to make sure.
By Anonymous on Tuesday, August 26, 2003 - 12:37 pm:
Smaller samples will equilibrate more quickly than larger samples (100µg verses 1g)
Water samples will equilibrate faster than organic solvent samples (looking for organic analytes)
Small vials are quicker than large vials.
Buffers/salts and HS mixers speed up equilibration.
A minimum of 5 min to a max of 60min are common values.
From 100 µL of Water I found 10-20min is best.
By Anonymous on Wednesday, August 27, 2003 - 03:07 am:
Have you thought about using other techniques like SPME,SPE or even solvent extraction?
I assume your matrix is water/acid (beverages?).
With headspace or SPME using sodium chloride is a good idea.
By Matthew on Thursday, September 4, 2003 - 12:15 pm:
Thanks for the tips "Anonymous" - they will come in handy.
And thanks for the suggestion David, but I have an HP automatic HS sampler so it was easier to use HS. I do have the manual SPME sampler, but I think there would be too much operator error doing manual sampling. My matrix was mostly water.
In the end, I used the increment function to increase the time 5 mins over replicate samples. The response was pretty much flat from 15 to 60 mins, so I guess 15 would have worked, but I used 20 to be sure. For 9 injections (stds and samples) I got a 2.9% RSD for the ISTD, so I figured it was at equilibrium.
Thanks again everyone,
By Anonymous on Friday, September 26, 2003 - 03:03 am:
I am working on a HS-methode to determinate flavors too. Matthew, how is your repetability with the HS-autosampler? My so not that good. Has anyone a good tip how to make the methode better, and witch are the parameters to lock at, specially?
By Rod on Friday, September 26, 2003 - 05:59 am:
From Anonymous on Tuesday, August 26, 2003 - above
The most critical factor in repeatablility is the proper crimping of vials.
The second is the proper choice of vial septa.
The third is sample preparation (if poor it becomes #1 factor).
If you have poor repeatability you must consider the reactivity of your compound, the inertness of your sample path, connecting lines, columns, valves, sample loops, etc.
You must understand the workings of the technique to get good results.
To see what can be achieved with a minimum of material look at Analytical Chemistry Vol 69, #11, p2221-2223.
With proper preparation, 50 ppm of analytes in a 1 mg of sample (25µL of solution) were measured with an RSD of less than 5% with linearities of 0.999+ with recoveries based on the regression line greater than 94%. The only exception was PYRIDINE, a very basic analyte (88%), and not an easy analyte in headspace.
I have found that headspace is only as accurate as the operator's skills in sample preparation, ie the operator is the limiting factor, not the technique.
I have even done triethylammonium salts (measured as the free base) with accuracy and linearity, albeit at levels 50 times higher than these noted above.
The above discusses volatile components, not semivolatile ones like phenol, dodecane and higher, etc.
Use SPME for semivolatiles.
The things NOT to do?
Put more than 1 mL of sample into a headspace vial and expect good RSDs in a timely fashion.
Heat reactive analytes for long periods of time.
Use aluminium faced septa with chlorides or chlorinated analytes.
Not use a stirrer or dissolve your sample completely.
Allow your samples to stand for hours before analysis.
Use bare rubber septa without teflon coating.
Use a crimper that does not give good seals.
Use metal transfer lines without a glass lining or equivalent.
Inject too large or too small a sample volume.
Use split injection (although it can work well if set up properly)
Generally, if you heat a sample for 5 minutes or more with less than 100µL of solution you will get REPRODUCIBLE results and LINEAR results.
You may be below the maximum recovery by 20-30%, but that may not be necessary to measure accurately the content of your sample.
Email me if you have questions in particular.