Chromatography Forum

We're trying to figure out the purity of butanol using GC-FID. We are using a SRI 6810 GC splitless / on column injection with 10 mL/min nitrogen carrier gas flow and uses Peaksimple software. The column we are using is a Restek MXT-WAX Polyethylene glycol nominal dimensions: 30.0 m x 530 µm x 1.00 µm.

We are injecting 0.1 uL of neat butanol and it overloads the detector with broad / sharkfin peaks for butanol. We want to do % analysis of other volatile components based on their area vs total area. What is the best course of action? We've tried different FID gains with little to no change in peak shape, or the peak gets cut off.

Is it possible to dilute the samples with water and still have good purity analysis using area / total area? If our current GC is not capable, what are some types or configuartions that would allow us to do this type of analysis?

Thanks in advance.

The sharksfin peak shape is generated by overloading the column, not by overloading the detector. That is why changing detector settings makes no difference.

GC-FID and comparing peak areas on one run is generally a poor way to determine purity unless the material is very impure. With pure materials either the main peak is above the linear range or the the minor peaks are below it. There is also a good chance that significant impurities either do not reach the detector, or are invisible when they do - water is the most troublesome of these.

Peter

You should be running split injection with a big split ratio not splitless/on-column.

Impurity testing of other alcohols in a "pure" alcohol is very tough although this does depend on how low an purity concentration you are looking for ?

Ideally you need to split the sample so you can see your main peak on scale. Then you can see what impurity levels you can see to determine.
I would not advise a 0.1ul injection as even from a 1ul plunger in needle syringe this is difficult to introduce reproducably.

GBLJoeTruong,

A peek at ASTM D5501 might be appropriate. Uses a big split, long column, FID detector and some interesting math in the calibration. Used to measure impurities in denatured alcohol headed for blending into gasoline.

Best regards,

AICMM

Thanks for all your help guys, really appreciate it.

When I do raw material purity checks I inject 0.2 ul (smallest the autosampler allows) with a 500:1 split. I still get huge peaks but I can get a good picture of the impurities.