Hi there.. I would like to seek for an answer regarding the method of silylation using MSTFA and pyridine. As I've go through many journal and protocol, the minimum weight of sample to be silylated is 1 mg (= 1000 ppm). But my problem is that, our GC-MS instrument are allowed to be inject with a sample at 1 ppm. I've done the serial dilution and dried the sample. But what I concern about is there any left compound after I've dried (I dry under fume hood) and filtered it. The results doesn't showed any peak except for the pyridine. Should I used 1000 ppm or 1 ppm? At what ratio of sample (g): MSTFA: pyridine should I used? Is it ok to inject 1000 ppm in GC column (BP5MS column)?
The MSTFA is no problem on the GC-column, more Problem is the GC-MS. The best is you switch off the ion-source until the MSTFA is eluted. I often use BSTFA 5000ppm without Pyridin which give reproducible results. My opinion is, the reaction takes place in the heated injector and not at roomtemp. during drying. The excess of derivating agent is essential for reproducibility.
Manfred, you highlighted what I consider to be a serious design fault in all commercial capillary GC-MS instruments. They need a means of dumping the injection solvent to waste, and for good reasons people don't often want a solution that has capillary connectors in the flow-path.
I wrote a note on RG about that: "Using a jet separator with shutter vane to divert the injection solvent in gas chromatography mass spectrometry (GC-MS)."
DOI: 10.13140/RG.2.1.2721.5848 2015-09-11
Since then, I reckon a better implementation would involve using the capillary column as the jet, simple pushing it away from the direction of the skimmer during injection so the solvent goes to the primary vacuum pump. Note that here the purpose of the jet separator is not analyte enrichment, which not everyone would need. It could act as a splitter though, particularly for hydrogen which is not efficiently pumped by high vacuum pumps.
I know this reply is a bit off topic, but think it worth circulating such undeveloped ideas because patent protection has, on the whole, tended to hinder progress in our field.
I have more experience with using BSTFA than MSTFA but have never had a problem with minimum amounts of sample needed to derivatize. I have used both on extremely small samples (much smaller than 1 mg - more in the few μg range) and injected minute amounts into the GC-MS using a very similar column. With BSTFA there is no need for pyridine; derivatization can be done in several organic solvents or with BSTFA alone (hexane usually works quite well). I evaporate off the excess derivatizing agent, whichever I'm using. It doesn't damage the GC-MS column, but it will produce a huge peak at the beginning of your chromatogram, as will any solvent you use to re-dissolve the sample for injection. I dry down under a gentle stream of dry nitrogen with very gentle heat (30-40°C) - whether you want to do this will depend on what your samples contain. The way to get round the huge peak at the beginning of your chromatogram is to set a solvent delay on the GC-MS, which you can do through the software. If you go into your method there is usually an option to set a delay. This automatically allows the solvent/derivatizing agent to pass through the column before the mass spec starts collecting data. Ours is set to 6 minutes but you may have to experiment a bit to get the time exactly right for your set up. By the way, did you zoom in to your chromatograms which contain the huge pyridine peak to see if any small peaks were eluting further on? Sometimes the solvent peak is so large that there appear to be no others until you zoom in. Your questions about concentrations are fairly meaningless without knowing more about your samples and your column (diameter, stationary phase thickness etc.). You should not inject any concentration which will either more than fill your injector when it vaporizes or will overload the stationary phase. 1000 ppm sounds rather too much to me. If you dry your sample off before injecting you can then dilute it in whatever amount of solvent you like before injection provided the solution fits into the autosampler vial. Most normally you then inject 1 μl of that into the machine. If you need to know concentrations you can work those out for the whole sample provided you know how much material you started with.
Valerie, just a couple of additional points:
1) Some derivatives (e.g. of amides and carboxylic acids) decompose on contact with active sites and traces of water. In my (old...) experience it was better not to try to remove the reagent in such cases. By contrast, some TMS ethers (steroids etc) are quite stable.
2) Solid samples that are not soluble in the derivatising cocktail sometimes take a very long time to react. It's best to make a solution first, for which pyridine is or was popular, partly because it was thought to catalyse some silylations, and partly because it's the usual solvent for methoxyamine, which is used for the preliminary oximation of ketones and aldehydes.
Pyridine always contains a trace of water (it's quite hygroscopic) and it's a convenient reagent for hydrolysing lactones of 4- and 5-hydroxycarboxylic acids which tend to form during sample extraction.
Dear Manfred, Christopher and Valerie...
Thank you for the explanation above. I've done another trial using 10 ppm and 1 ppm. The result showed a presence of several sugars ( d-mannose, d-galactose, fructose, sucrose and etc.) but I'm not interested on that particular sugar. What I'm looking for is polyhydroxy alkaloid (or imino sugars) which contain nitrogen atom and OH. My sample is 100% methanolic extract. As I searched and go through several journal, they also used 50% aq. ethanol and 50% aq. methanol. Therefore, I'm gonna try this two aq. solution method and run on GC. If the result shows an imino sugars, then I know that my extraction is my main problem. But if I still get that fructose,galactose,mannose sugar meaning that there is problem with derivatization or GC.
Christopher: As far as I'm concern, they allowed us to inject below than 10 ppm concentration because there are afraid that the filament will burn if the concentration is too high.
Zuriah, if you try this reaction in aqueous solution I don't think it will work. Take one ml of your solution to dryness in a 2mL GC autosampler vial, then add 100 uL of MSTFA and 900 uL of Pyridine. Mix and allow it to react for about 30 minutes at 50C using a thermo mixer, or a simple heating block and re-shake it every 10 minutes or so. After it cools down, just inject it directly into the GC. I usually only inject 0.2 uL, this saves on detector fouling but you may need more depending on your detection limit requirements. MSTFA can form TMS ethers on sugars, but to find out exactly the distribution (ie x1 TMS, x2 TMS etcc) you should do some homework in the literature. I use this procedure for doing amino acid dervatizations on soil pore water samples and it works pretty good, capping the active amino hydrogens as well as the carboxylate hydrogens. Cheers, Joe.
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Column