GC FID parameters

[Bellyup]

Hi
Im running the same sample at different parameters and was wondering how do i determine which is the best peak to choose the right parameter.
Here is just changing the HS oven temp and nothing else.
Do i go by highest area i can obtain?

Note, this is one solvent peak, overlayed

[James_Ball]

Hi
Im running the same sample at different parameters and was wondering how do i determine which is the best peak to choose the right parameter.
Here is just changing the HS oven temp and nothing else.
Do i go by highest area i can obtain?

Note, this is one solvent peak, overlayed

I can't load images at work, but I would look at not only largest area, but also peak shapes. If a peak is fronting then it could be overloaded, or if tailing then something is holding on to it or possibly need higher temperatures in the transfer lines.

Adjust the HS oven temp for highest areas, then optimize other parameter if the peak shapes are bad. Just do one parameter at a time. It always takes several iterations to optimize a method with so many parameters.

[Bellyup]

Hi, thankyou!
The peak shapes are good, but if i graph peak area against height then i see that the height doesnt increase alongside peak area, it starts to level off while the area keeps increasing

[James_Ball]

Hi, thankyou!
The peak shapes are good, but if i graph peak area against height then i see that the height doesnt increase alongside peak area, it starts to level off while the area keeps increasing

Does the peak have a flat top? If so then it is exceeding the limit of the detector, and with larger areas it is just a little wider peak but not any higher.

If you display the signal on the front of the GC and it goes to maximum and stays there a few seconds before dropping it will definitely indicate an overloaded condition.

Have you tried a higher split ratio to see what it does?

[Bellyup]

Hi, thankyou!
The peak shapes are good, but if i graph peak area against height then i see that the height doesnt increase alongside peak area, it starts to level off while the area keeps increasing

Does the peak have a flat top? If so then it is exceeding the limit of the detector, and with larger areas it is just a little wider peak but not any higher.

If you display the signal on the front of the GC and it goes to maximum and stays there a few seconds before dropping it will definitely indicate an overloaded condition.

Have you tried a higher split ratio to see what it does?

I can see the peak tips on the software, but havent looked on the GC itself, its not flat on the software, ive been trying different split ratios, low splits ie 10:1 result in wide and leaning peaks, higher splits eg 90:1 are straight and narrower, the higher the split the less area i get, so im thinking of going for 40:1, nice shaped peak, still high peak area.
Ive gone up to a oven temp of 130c on the headspace sampler and increased the sample line and transfer line temps too, still seems like theres room to go higher but im hesitant

[James_Ball]

Hi, thankyou!
The peak shapes are good, but if i graph peak area against height then i see that the height doesnt increase alongside peak area, it starts to level off while the area keeps increasing

Does the peak have a flat top? If so then it is exceeding the limit of the detector, and with larger areas it is just a little wider peak but not any higher.

If you display the signal on the front of the GC and it goes to maximum and stays there a few seconds before dropping it will definitely indicate an overloaded condition.

Have you tried a higher split ratio to see what it does?

I can see the peak tips on the software, but havent looked on the GC itself, its not flat on the software, ive been trying different split ratios, low splits ie 10:1 result in wide and leaning peaks, higher splits eg 90:1 are straight and narrower, the higher the split the less area i get, so im thinking of going for 40:1, nice shaped peak, still high peak area.
Ive gone up to a oven temp of 130c on the headspace sampler and increased the sample line and transfer line temps too, still seems like theres room to go higher but im hesitant

You can adjust your split up until you find the lowest calibration standard has too little area/height and still be good. Especially if the samples are a little dirty, the more split you can run the longer the column will last.

If the low points become less linear then lower the split and get more on column, if the high points begin to fall over away from linear then increase the split, that is usually a good starting point to follow.

[Bellyup]

Does the peak have a flat top? If so then it is exceeding the limit of the detector, and with larger areas it is just a little wider peak but not any higher.

If you display the signal on the front of the GC and it goes to maximum and stays there a few seconds before dropping it will definitely indicate an overloaded condition.

Have you tried a higher split ratio to see what it does?

I can see the peak tips on the software, but havent looked on the GC itself, its not flat on the software, ive been trying different split ratios, low splits ie 10:1 result in wide and leaning peaks, higher splits eg 90:1 are straight and narrower, the higher the split the less area i get, so im thinking of going for 40:1, nice shaped peak, still high peak area.
Ive gone up to a oven temp of 130c on the headspace sampler and increased the sample line and transfer line temps too, still seems like theres room to go higher but im hesitant

You can adjust your split up until you find the lowest calibration standard has too little area/height and still be good. Especially if the samples are a little dirty, the more split you can run the longer the column will last.

If the low points become less linear then lower the split and get more on column, if the high points begin to fall over away from linear then increase the split, that is usually a good starting point to follow.

Thank you James, appreciate your help!