Retention Time

[PrithiRao]

I use the GC/FID . I analyze the % of solvents present in our formulation. When I run THF alone, I get the retention time as 2.868. But in the mixture( with other solvents) the retention time is shifted to 2.920. Does this make a huge difference? and why does this happen

[AICMM]

In light of your other post, what you have is a small amount in one case and a neat (or almost neat) in another? That being said, you can have RT shifts due to column overloading so that full concentration is at one RT and low concentration is at another.

Best regards,

AICMM

[PrithiRao]

Thank you. I use a split ratio of 20:1.
I also wanted to know if there is a method to understand the purity of a solvent using the FID. When the supplier says there is 300ppm of water present in the solvent, we usually check it with the karl fischer titrator.For a purity of 99.7%, can I get this kind of accuracy in the GC? I do understand that FID will not detect water. What I am doing is injecting a 0.2 micro liter of solvent as is and checking the purity as against the std which is 99.9% purity. The problem with this method is I keep getting varying results. Should I be increasing the split ratio?

[aldehyde]

What other solvents is it mixed with? And at what concentrations?

The analysis of a sample matrix, like you describe, has its own separation characteristics as compared to the injection of a single, neat compound. This is called a matrix effect, and can be caused by many different things.

For example, if your sample is condensed at the head of your column with a low starting oven temp the analyte's boiling point may be affected by the other analytes it is dissolved with. Or if you had a sample that contained reactive compounds they can undergo reactions, especially in a heated environment. If your sample contains non-volatiles that accumulate on the liner it basically becomes a reaction zone. The reacted analytes may have different retention times, and will likely come out as a smeary tail rather than a nice sharp peak.

I don't know if your method for checking solvent purity will work, you should keep in mind that there is some error associated with the injection (hopefully around 0.3-1%.)

Your split ratio is probably OK but check out this calculator which helps design proper inlet conditions: http://www.chem.agilent.com/en-US/Techn ... ev205.aspx

[chromatographer1]

While in LC the purity of a sample is compared against a known purity standard, GC doesn't work well when you are trying to compare a balance against another balance.

With GC usually the impurities are measured and the balance is given a purity by calculating it by difference. For example, if a solvent is known to only have water as an impurity then its purity is determined after only a water titration. I have tested Acetonitrile in this manner, although I tested for water by GC, using a direct injection, a Chaney syringe, and used a standard addition method to determine the solvent has 0.27% water, which met the specifications of less than 0.3 % water. No other peaks were detected by high resolution capillary GC.

Using a split will usually cause a variation in peak size of 1% RSD, and this does not account for a sample volume variability by manual syringe which can be 2% or more.

best wishes,

Rod

[dblux_]

Thank you. I use a split ratio of 20:1.
I also wanted to know if there is a method to understand the purity of a solvent using the FID. When the supplier says there is 300ppm of water present in the solvent, we usually check it with the karl fischer titrator.

You will get different results with different methods (technics) employed. Follow standards, if established. K-F titration is the best in your case I think.

... What I am doing is injecting a 0.2 micro liter of solvent as is and checking the purity as against the std which is 99.9% purity. The problem with this method is I keep getting varying results. Should I be increasing the split ratio?

Do you get reproducible results doing 0.2 µL injections (really doubt it)?
You won't get right answer re. split ratio unless you tell us what is your analytical column. You know why.

[PrithiRao]

This question is in regard with just testing a purity of a solvent like THF, Acetone etc( we are more concerned about moisture), not in any matrix .But since I cannot detect water I was more trying to see if I could results like 99.7 % for purity but yeah its not reproducible with 0.2microliter or with 0.1micro lite injection.
I was just wondering if there was some other method for this. I have not done a standard addition method before so not sure how to go about it.

[aldehyde]

I'm not a big fan of 0.2 uL injections, I prefer 0.5 or 1 uL with a split if needed. I routinely get RSD% of 0.3-0.8% when using an autosampler and split inlet conditions.

[Peter Apps]

To avoid overloading the column, and going above the linear range of the detector you would have to inject a nearly pure solvent with a high split ratio. The repeatability of these injections is too poor to distinguish between 99/7% and 100% unles you do several replicate analyses of the sample and the standard.

Peter

[JMcK]

PrithiRao, you have to be careful to distinguish purity from assay. Usually a vender will report GC purity; and that will indicate the area percent observed in GC. It will not account for the analytes that are not observed by GC. Assay accounts for all impurities present. This can be difficult to calculate because you need to know, or estimate, response factors for all the compontents.