Ether analysis

[washington]

We just had to change the range of our curve for the analysis of Ether ( new range 0.01 to 5.5 ug/mL) and are having a hard time reproducing a linear curve. Is ether linear over this range or do we need to break the curve up into two separate ones?

[chromatographer1]

Ether response is linear if you are using an FID. You should determine why the non-linearity occurs and see if it can be eliminated.

[washington]

Thanks for replying. We do run FID. Being new to GC, what are some of the pitfalls I should be concerned with in running ether analysis?

[chromatographer1]

You have not told me anything about the matrix and how you are doing your sample prep.

I don't know how I can give you any meaningful information if I don't know the particulars of your application. I am not nosey but I don't care to write a book describing all possible issues with ether from any and all matrices from which you could possibly be working.

If you do not care to document on this forum the application details you may email me directly and I would be glad to share any experience or knowledge I can.

rgeorge @ sial . com (omit spaces)

[washington]

We run a Varian CP3800 GC with a 1079 universal liquid injector, split less. Gasses are Helium, Hydrogen and Air with FID detector. No sample prep. Inj. temp. 225°C, column DB-WAX
col. oven
temp rate hold total
°C °C/min. min. min.
60 0 0.5 0.5
90 20 0.5 2.5
170 50 2.5 6.6

[chromatographer1]

It appears that you are measuring ether in air or some other matrix without any sample prep at a concentration of 10 to 5500 ng per milliliter of sample volume.

I have to assume you are not injecting a large volume or a very tiny volume. I also have to assume you are not creating a coelution from any decomposition of your matrix, whatever that is.

I have to assume you do not have a leak in your injection system and the vials in which you hold your sample if it is liquid are not leaking or undergoing some temperature change from their environment.

I have measured ether from 0.5ng to 1000ng per injection on a FID (a Varian 3500 by the way) where the samples were correctly prepared and properly sealed and found a linearity of greater than 0.99985 (yes 3-9s and an 85) with a recovery of 97% at a 50ng level compared to the regression line.

Linearity studies were conducted to 5,000 ng and gave the same slope for the calibration curve.

I have to assume if you are not producing a linear response you are somehow allowing your ether to escape before it reaches the detector or you are not preparing your standards correctly.

Review your handling of your sample and the means of which you are injecting the sample and determine where the loss of ether occurs.

Have you confirmed you can produce linearity by using an internal std that is non reactive and relatively non-volatile?

you may email me directly or call if you have any particular question to ask off line.

best wishes

rgeorge @ sial.com
814-359-5737