Problems with Sample prepartion

[ridge]

Hi Everyone,

I am having problems in preparation of sample prep for ethanol analysis using n-propanol as the internal standard. I use small vials in which I add 2.5 % v/v internal standard solution and then the unknown ethanol concentration using micropipettes to maintain accuracy. To test for the concentration of internal standard, I make a sample prep of known ethanol concentration such as 100% v/v and then try to analyze in the GC.
My results from GC analysis varies from one sample prep to another although I use the same internal standard and maintain similar procedures for sample preparation. Even for known ethanol concentration, it takes many attempts in preparing appropriate sample prep to achieve the required concentration.
Since I have less than 0.5% RSD for my injections, I presumed the GC is functioning well.
If anyone can suggest some comments on improving the sample preparation technique it could be very helpful.

Thanks in advance,

RD

[Consumer Products Guy]

We add the sample to a 20 or 40 ml screw thread vial, then (quickly) add accurately a measured amount of water already containing the n-propanol internal standard, cap, shake, transfer a portion to an autosampler vial, and inject 0.5 ul. We get good precision.

[WK]

Ridge,
Look through the archives since something similar has been discussed before.
Are you injecting 100% ethanol (with a little n-propanol) - perhaps you are overloading? Maybe your n-propanol concentration needs to be closer to the ethanol in samples and standards.
Are you using this 100% as a single point calibration? If not how do you go about determining the %v/v ethanol in the standards you have created - density and alcohol tables??
WK

[ridge]

Thanks for the suggestions.
I actually use a method developed earlier. In this method i have ratios of concentration on x-axis and ratios of area count on y-axis. I use the equation of the line (r2=.99) and then calcuate the unknown concentration. Since the results are varying from one sample prep to another, i was suspecting the internal standard solution. I have sometimes ended up making many internal standard solution (2.5 % v/v) in trying to get accurate concentrations of the known samples .Again any suggestions in improving consistency in sample preparation would be very helpful...

thanks,
ridge