Chromatography Forum

I need to determine glycerin purity via GC. My sample is crude glycerin from the biodiesel process. The glycerin is typically 75-85% pure, 5-10% h20, 5% salts (soaps, sodium sulfate, sodium citrate) and the rest MONG (material organic non glycerol).

I have one GC with on column injection, FID, 30m DB5-HT with guard column. Also one GC with SSL, FID, 15m Stabiliwax (PEG) column also with guard attached.

My thinking is to use the DB5-HT column and derivitize with MSTFA first, then dilute with methanol. Make a calibration curve using reagent grade glycerin. I guess I would first dry the unknown glycerin sample, then derivitize, but I am worried about gunking up the on column GC due to salts.

Is there anyway I could use the PEG column with SSL? I would rather get the junk on the liner than on the column, but I am worried about too much moisture being a problem.

I would also rather not use MSTFA, could I just turn up the inlet temp on the SSL to above the boiling pt of glycerin or would it decompose? Thanks any input is appreciated

What is SSL ?

You should NEVER dilute a silylating reagent with methanol.

Pentane or acetonitrile !

This is a case for a packed column analysis with a guard column or at a minimum use a drop down packing at the injector.

Second choice? Do not use an on-column injection.

I would suggest a glycerol per weight would be the best method, and use an internal standard in the reagent for both sample and calibration standard reference preparation. A lower concentration (1mg/mL) preparation (glycerol in reagent) comparison would be a good idea

best wishes,

Rod

What is SSL ?
...
Second choice? Do not use an on-column injection.
...

Seems he considers using split/splitless injector.

I recommend DMF or pyridine as solvent, and to make trimethylsilyl derivatives; use your DB5-HT column. We use BSTFA, just add to our DMF or pyridine solution, shake, and inject. Yes, youneed excess derivatizing agent. But your ratio of analyte to H2O is very high, you should be able to make standard and sample of 0.1 g per 100 ml solvent or less. Then add portion to some derivatizing agent.

If you really want to try as underivatized, the best column I've experienced for underivatized glycerin peak shape is a -624 type column; see the USP glycerin monograph.

Thanks for the input!

I was planning on derivitizing the glycerol with mstfa, let sit for 15-30 min, then adding the methanol/solvent. I was assuming any leftover mstfa would react with the methanol and not cause a problem on the column. However it seems like what is normally done is to add the solvent then add the MSTFA...in which case methanol would not work, I do have pyridine on hand so that shouldn't be a problem.

What if the glycerol is properly dried of h20, and derivitized in pyridine. Could I inject onto the PEG column or would there still be damage to the stationary phase?

I would really like to use the split splitless injector as I have liners with a glass wool plug, that would take care of the salts/contaminants, i guess I could swap out the PEG column for the DB5 if I really had to...

The excess derivatizing chemical will not damage the DB5-HT phase, that's why nobody tries to "react" the excess. You do not need to dry off the water in the sample (you will lose some glycerin if you try that), the water will also react with the derivatizing chemical, OK. we assay stuff up to 90% water using this derivatizing technique.

However, the derivatizing chemical will DEFINITELY ruin a PEG column !

CPG wrote:

"However, the derivatizing chemical will DEFINITELY ruin a PEG column ! "

He is absolutely correct. BY NO MEANS should you EVER inject a silylating reagent on a PEG column.

You can use silylating reagents on 100% Methyl silicone, 5% phenyl silicone, 20% phenyl silicone, or 50% phenyl silicone, Fluoro propyl methyl silicone, or Apiezon grease phase columns.

You should not use methanol or any alcohol as a solvent as not only will you consume the excess reagent, you will also enter into an equilibration reaction with the derivatized glycerol, which can easily be pushed toward the free glycerol side of the equation.

Stay with solvents which assist in forming the silyl ether, such as DMF, pyridine, acetonitrile, etc. to make any desired dilutions. But using 100% MSTFA is the best option for avoiding interferences with the TMS-glycerol peak.

best wishes,

Rod

Good stuff! Thanks guys!