Chromatography Forum

My lab is developing an OVI method for various simple sugar products using FET. Given the elimination of a matric effect by this technique, is it scientifically valid and compliant to apply the validation data/result from one sugar to the other sugars? The final method would prepare all samples similarly.

regards..

How on earth do you get full evaporation of a sugar ?

Peter

He must mean an aqueous or other solvent (DMSO?) sugar solution. The whole idea of FET is that there is minimal matrix effects so the type and ammount of sugars from sample to sample should have no significant effect on results.

He must mean an aqueous or other solvent (DMSO?) sugar solution. The whole idea of FET is that there is minimal matrix effects so the type and ammount of sugars from sample to sample should have no significant effect on results.

Could be, but when the solvent has been evaporated he will still have a layer of syrup in the bottom of the vial.

Peter

possibly depends on how much he diluted the sample. Since with FET you can only put a small amount of sample into the vial at most there will be some crystals at the bottom. I ended up going with FET for my residual ethanol method and am working on a diacetyl method. Most of my samples are also carbohydrate rich (they are starch encapsulated). I take 1 g sample to 10ml water with acetonitrile as my ITSD. I then put 5ul into a 9ml vial on my Tekmar 7000 and heat it to 120 deg C. LOQ is 10ppm in the vial using GC/MS SIM so with samples 100ppm.

possibly depends on how much he diluted the sample. Since with FET you can only put a small amount of sample into the vial at most there will be some crystals at the bottom. I ended up going with FET for my residual ethanol method and am working on a diacetyl method. Most of my samples are also carbohydrate rich (they are starch encapsulated). I take 1 g sample to 10ml water with acetonitrile as my ITSD. I then put 5ul into a 9ml vial on my Tekmar 7000 and heat it to 120 deg C. LOQ is 10ppm in the vial using GC/MS SIM so with samples 100ppm.

Impressive - at the LOQ you only have 5 ng of your analyte in the vial, so about 0.5 ng on the column if you are taking 1 ml headspace and transferring splitless.

Peter

Actually I use a 5:1 split.

1 ml sample loop 1.3 ml/min column flow with septum purge total flow down the transferline is ~10 ml/min the minimum I can get away with. I tried a pulsed split at 3:1 with 3ml/min column flow but the 6890 EPC couldn't generate it repeatably.
5ul is the most I can put into the vial according to PV=nRT

PV/RT =n ((1atm)(.0092L))/((.08206)(393K)) n=.000285 moles *18 g/mole= .0051 g =5.1ul (1g/ml)

I am wanting to get a 2ml loop to boost sensitivity. at 10ml/min that is sufficient to sweep the loop 5 times under a 1 min injection.

Actually you can go up to 7ul if you allow a static head pressure up to 1.5 atmospheres 7.35 psi. Beyond that you are pushing it. Kolb also makes the point of the moisture already in the vial from ambient conditions.

I ran it today and got excelent linearity and sensitivity doing diacetyl with acetonitrile as an internal standard on a typical Innowax 30m .25mm .25um film column (1.3 ml/min) at 40 deg C diacetyl came off right before acetonitrile at about 4.5 min. I could probably go even lower than 10 ppm.

SIM diacetyl 43 & 86 quant 43
41 for acetonitrile.

Same procedure for ethanol only 45/46 m/z

I upped it to 7ul in a 9ml vial 9.2ml nominal volume. beyond that you are probably risking overpressure.


Full evap works much better than standard headspace on my samples and since diaceyl and ethanol have such high partition coefficients (k) in water you aren't giving up much sensitivity.