Variation in results

[catrinaf]

I have an Agilent 7820a with headspace sampler. I have carried out a run on the same sample from the same vial, using the same method (basically everything is exactly the same for each run) about 10 times, straight after each other. I have noticed that after each run the given result is decreasing (both area under peak and height of peak have decreased). The sample I am testing is a hydrocarbon and solvent solution, and I am looking at the hydrocarbon result. The peak I am using is very small (0.3%)and over the 10 runs this has dropped to <0.2% (quite significant for our purpose) Any body have any suggestions - I could understand it if the result was increasing (due to the volatile solvent used), but not decreasing

[kubowicz.tomasz]

Hello

If you're using the same vial all the time there is different coefficient (partition) ratio between liquid and gas phase for your analyte. That is why you have smaller peaks. Basically there are different thermodynamic parameters in the vial.
Also if vial is pierced first time it could be leaking with next injections.

Regards

Tomasz Kubowicz

[catrinaf]

I don't usually use the same vial all the time, I only used it for these 10 runs to see if the result was repeatable (which it isnt)

[Peter Apps]

Welcome to the forum. Each time you take a sample from the vial the amount of the target analyte in the vial decreases, so the area of the analyte peak should get smaller and smaller.

It is not common to use headspace to analyse minor components of solvents - is there a reason why you do not do conventional split or splitless injections ?

It would be a big help if you provided details of your instrument and conditions, and what it is you are trying to analyse. Then we will not have to troubleshoot by guessing.

Peter

[catrinaf]

The solvent is only used as a diluent, and is not what is being analysed.

I don't understand why you say " Each time you take a sample from the vial the amount of the target analyte in the vial decreases, so the area of the analyte peak should get smaller and smaller" - the amount injected is the same every time (0.5µl) so shouldn't the peaks be the same size?

I've not must experience with GCs, so my use of the word "headspace" may be incorrect.

We use an auto sampler to inject 0.5µl sample (the sample is a methacrylate diluted with acetone), we do use a split injection

Not sure if this helps any

[rb6banjo]

The partition coefficient stays constant for a given set of conditions. It's a thermodynamic parameter. What you're changing is the amount of analyte that is actually in the vial.

In reality, what you're actually performing a technique call multiple headspace extraction (MHE). Kolb and Ettre describe the mathematics of this in detail in their book "Static Head-Space Gas Chromatography", Wiley is the publisher. If you plot the logarithm of the peak area for you analyte (y-axis) vs. the extraction number (# of times you analyze the vial), the shape of the curve can tell you a lot about what type of system you have (partition, adsorption, etc.). Basically, in the end, if you can get a detector response for a known mass of analyte, you can use the MHE data to extrapolate the mass of that analyte in the sample - and as a result the concentration of your analyte in the sample. Check it out. You may find it useful.

[catrinaf]

Thanks rb6banjo - if I could understand it, I would check it out

[rb6banjo]

I can't go into all of the details here but if you could get your hands on the book and read their description, I think you could understand it. It's pretty cool actually.

[catrinaf]

Will definitely take a look thank you

[Peter Apps]

The solvent is only used as a diluent, and is not what is being analysed. You are analysing acetone for its content of methacrylate. The methacrylate is the analyte, the acetone solution of methacrylate is the sample

I don't understand why you say " Each time you take a sample from the vial the amount of the target analyte in the vial decreases, so the area of the analyte peak should get smaller and smaller" - the amount injected is the same every time (0.5µl) so shouldn't the peaks be the same size? the peak that you see on the chromatogram is the detector's response to analyte that has been injected to the GC and passed through the column. That analyte came out of the sample vial. By taking analyte out of the sample vial you have reduced the quantity of analyte in the vial. In headspace analysis the peak size is proportional to the quantity of analyte in the sample vial, since you now have less analyte in the vial, the next peak will be smaller, and so on. NB that this depends on headspace conditions - a 0.5 ul sample volume suggests that you are injecting a liquid sample

I've not must experience with GCs, so my use of the word "headspace" may be incorrect. Headspace is the gas phase in a vial above the sample - headspace analysis depends on the composition of the headspace being a function of the composition of the sample. As above it would be very unusual to take a 0.5 ul headspace sample, and I suspect that you are doing a liquid injection. Does the autosampler syringe remove liquid form the sample vial ?

We use an auto sampler to inject 0.5µl sample (the sample is a methacrylate diluted with acetone), we do use a split injection

Not sure if this helps any

It does, you are not doing headspace analysis, you are doing a conventional liquid injection. With a liquid injection analyte peak area should stay the same from run to run

The first troubleshooting step is to make up some sample solution and divide it between five vials. Run a sequence of those 5 and look at peak areas.

We still need to know your instrument set up and operating conditions. The more you tell us the more we can help.

Peter

[catrinaf]

I have run the 5 samples as you suggested and the areas I got were:1074670 (0.245%), 1044646 (0.239%),1356364 (0.31%), 936956 (0.214%), 1177005 (0.269%). The area we should see is 0.3%

Not sure what information you require about the set up but here is some info from the method, let me know what else you need:

Injection volume - 0.5µl

Back inlet
Pressure - 15.008psi
Total Flow - 22.848ml/min
Septum purge flow - 3.348 ml/min
Split ratio - 2:1
Split flow: 13ml/min

Column - HP5

Back detector FID
H2: 30ml/min
Air flow: 400ml/min
Makeup flow: 25ml/min


10.4 min run - starts at 90°C, then 25°C/min to 250°C for 4 mins

[kubowicz.tomasz]

Hello

I think you have problem with:
-syringe - check if it is tight enough (perhaps it is leaking) or put new one in.
-inlet - you can have leak. Make sure all connections are ok (septa, spit vent trap etc)

Regards

Tomasz Kubowicz

[Peter Apps]

I have run the 5 samples as you suggested and the areas I got were:1074670 (0.245%), 1044646 (0.239%),1356364 (0.31%), 936956 (0.214%), 1177005 (0.269%). The area we should see is 0.3%

Not sure what information you require about the set up but here is some info from the method, let me know what else you need:

Injection volume - 0.5µl

Back inlet
Pressure - 15.008psi
Total Flow - 22.848ml/min
Septum purge flow - 3.348 ml/min
Split ratio - 2:1
Split flow: 13ml/min

Column - HP5

Back detector FID
H2: 30ml/min
Air flow: 400ml/min
Makeup flow: 25ml/min


10.4 min run - starts at 90°C, then 25°C/min to 250°C for 4 mins

Thanks. At least there is no trend as there was with repeat injections from one vial.

Your split ratio is too low - the gas controls struggle to maintain low flow rates. Try increasing it to 10:1, or use a proper splitless injection with the split closed for 30s.

The column temperature programme rate is very high, why such a high rate ?

What size (volume in microlitres) syringe are you using. Repeatability gets markedly worse for injections of less than 10% of maximum volume. Is there a specific reason to inject 0.5 ul rather than 1 ul ?

What is your inlet temperature ?, and what are you column dimensions (length, i.d., film thickness) ?. What kind on liner do you have in the inlet ?.

The acetone solvent peak might be above the linear range of the detector - is the peak flat or rounded on top if you adjust the vertical scale to show the whole peak ? How repeatable are the areas of the acetone peak and the analyte peak in your set of five runs?

How do you know that the area ratio should be 0.3 % ?

Peter

[catrinaf]

Thanks. At least there is no trend as there was with repeat injections from one vial.

Your split ratio is too low - the gas controls struggle to maintain low flow rates. Try increasing it to 10:1, or use a proper splitless injection with the split closed for 30s. Thank you, I will try this to see what effect it has

The column temperature programme rate is very high, why such a high rate ? The peak i need doesn’t show until over 200°C, and we need a quick turnaround time

What size (volume in microlitres) syringe are you using. Repeatability gets markedly worse for injections of less than 10% of maximum volume. Is there a specific reason to inject 0.5 ul rather than 1 ul ? we use a 10ul syringe – its 0.5ul because that was the method in place when I joined the company, but if it makes a difference its definitely something i will look at changing

What is your inlet temperature ?, and what are you column dimensions (length, i.d., film thickness) ?. What kind on liner do you have in the inlet ?. inlet temp is 300C (i think) – column is 30m x 320um x 0.25um, liner is split, single taper, glass wool, deactivated, low pressure drop,

The acetone solvent peak might be above the linear range of the detector - is the peak flat or rounded on top if you adjust the vertical scale to show the whole peak ? How repeatable are the areas of the acetone peak and the analyte peak in your set of five runs? Acetone peak is flat, the method is not set up to show the acetone peak area, as this is just used as a diluent, so not interested in the area

How do you know that the area ratio should be 0.3 % ? made up the sample myself

[Peter Apps]

"The acetone solvent peak might be above the linear range of the detector - is the peak flat or rounded on top if you adjust the vertical scale to show the whole peak ? How repeatable are the areas of the acetone peak and the analyte peak in your set of five runs? Acetone peak is flat, the method is not set up to show the acetone peak area, as this is just used as a diluent, so not interested in the area

How do you know that the area ratio should be 0.3 % ? made up the sample myself"

Can you provide more details of what you are analysing ? To calculate the percentage results you must be dividing your analyte peak area by the area of another peak, yes ? What is that other peak ?, are you adding an internal standard ?

Peter