Phenol Disapearance

[Ricardo]

Hi all. I am analyzing phenol in water from 100 to 2000 ppm using A DB-5 capillary column, split injection to an FID. My instrument is a HP 5890. with auto-injector.
After external calibration and a couple of sample runs, my phenol peak has disappeared (std and sample). I have changed the glass liner, septum, cut the column, baked out the gc, checked all flows and temperatures/settings, reduced the split. I have run other test analytes such as acetone and ethanol and the chromatography is great.
I am now having the student mix a 1 % test sample.

My temperatures are: injector 175, column 90 isothermal for 6 minutes, detector 250, injection volume 0.5 ul. The phenol was tailing quite a bit as well.

I am lost on this one. Help please!

Ricardo

[Consumer Products Guy]

Phenolics can be tricky because of the phenolic group acting acidic. Personally, I'd use HPLC with an acidic eluent; when we do phenolics by GC here, we derivatize to the more-stable trimethylsilyl derivatives.

[chromatographer1]

The problem is your matrix. Water does nothing good for silicone phases and fused silica surfaces.

Combined with other components from your mix a lot of chemistry can be going on.

Trace amounts of phenol can disappear easily. (tailing is an indication of adsorption)

This could happen with any company's columns.

CPG is quite right in his comments. Use the proper tool for the analysis at hand. That is why we are paid the big bucks.

best wishes,

Rod

[Peter Apps]

Assuming that you are stuck with GC (not all labs have every instrument in the catalogues just in case a tricky analysis comes up !) your first step needs to be to get rid of the water. Acidify your sample to PH less than 2, and extract with organic solvent. Inject that instead of water. Sadly, you have probably destroyed the column already, you would be better off with a wax phase anyway. Increase the inelt temperature - the phenol will stick to the liner and it needs some heat to shift it.

Good luck

Peter

[AICMM]

Ricardo,

As per Peter, a solvent extraction is definately called for. Then, get very meticulous about clean liners and column trimming and you should be able to do these levels reasonably well. I think the DB-5 is probably a good column for this app (when using solvent injections) but you will need to trim regularly. You may or may not want wool but if you do try wool, an absolute minimum, a wisp at most.

Best regards.