I am using GC to analyze a sample of phenol/water at about 75/25 percent. The goal is to consistently determine how much phenol exactly is present in the sample. Water is determined by Karl Fisher and the total should be close to 100%. GC results using a megabore column, split injection (single taper liner, 1uL volume) are poor, yielding phenol results that vary widely and are typically low by as much as 15%. The sample usually can be made homogeneous and is then dissolved in acetone about 0.05g/25mL.

Not knowing the source of your samples, I don't know if this is possible but... Could you have phenol salts in your samples? Does any part of your preparation include acidification of the sample?

I second the motion by Anonymous. You need to adjust the pH low to make sure you don't form phenolic salts.

You will also need an internal standard technic, or a really good autosampler. Normal human beings cannot inject 1 microliter with much better than a +/- 5% repeatability on a good day.

Silanize everything that touches the sample too.

Good Luck

Why don't you do this by HPLC? Use a C18 column and methanol/water or acetonitrile/water. I don't think you will have any problems and you won't need to get entagled with all the difficulties of injecting water on to a capillary GC column.

Thanks Bill Jenko and Anonymous. I had not thought of pH adjustment and will investigate. Small Bore - your suggestion is exactly how I would handle this except the project leaders committed to buying a GC without consulting Analytical personnel. Typical that we have to clean up their mess. Any one have thoughts on how this type of sample would behave in the GC injection port? Descrimination?

Use a very inert liner, I would recommend the Restek Siltek liners. If you run splitless, there may be a problem with the solvent expansion volume, depending on the volume of the liner used. Use a fairly low injection port temperature to minimize solvent expansion.