MSTFA product stability

[joanne_y]

Hello,
How stable are MSTFA products after sitting at at ~20 °C for ~4-6 hrs prior to analysis?

I prepared 6 samples from tissue extracts, derivatized with 40 uL MOX/ pyridine followed by 40 uL MSTFA. I loaded the samples onto the autosampler. The first 2 samples looked good - spiked standards were derivatized. The last 4 samples did not appear to be derivatized. There was no trace of the standard (pentadecanoic acid) or of any derivatized peaks. I added another 40 uL MSTFA to the 4 samples that did not work and was able to "rescue" the samples - the standard was derivatized and the normal peaks corresponding to sugars, fatty acids, fatty alcohols were apparent.

I am using screw top PTFE caps. From the literature, I would expect that products might degrade after 24 hrs but perhaps the timescale is more like 4-6 hrs? If so, should I only run 2 samples at a time and store the remainder at -20°C prior to analysis?

I'd greatly appreciate any input or tips. Thank you!

[Nathan Hawkins]

MSTFA samples are generally stable for 48 hours provided the samples and reagents are dry (though batches sizes are best limited to 24-hours).
1. Keep MOX solid (it's hygroscopic) in a desiccator and prepare freshly for each batch of samples
2. Keep Pyridine in a desiccator, purchase in small quantities (100mL) and use within 6 months. Best to remove the crown cap and keep the bottle lid screwed tight (if you don't remove the crown cap, as soon as you pierce it, you'll extract the seal into the solvent)
3. buy MSTFA (with 1% TMCS) in ampoules
4. dry you samples thoroughly in a centrifugal vacuum concentrator (overnight) or Turbovap
5. use PTFE backed silicone or red rubber vial caps (pure PTFE discs won't seal)
6. if you are able to, mix (750rpm) while heating [to aid redissolution of dried residue]
the fact that you were able to 'rescue' the samples with another 40µL MSTFA suggests that the reagents or samples weren't sufficiently dry (adding more, provided an excess) - note TMCS does act as a catalyst, it also scavenges moisture

TMS derivatives are equilibria - there will always be some cleavage that starts as soon as you take the sample out of the incubator (if you're preparing in batches, rather than doing just-in-time derivatization with a robotic sampler, a cooled sample tray will to 10 °C will improve stability and reduce intra-batch variance).

While a lot of labs do both MOX and TMS derivatization at the same temperature and in equal ratios, Thomas Moritz and John Halket independently did work on MOX-TMS, showing 30 °C/37 °C were optimal for MOX and TMS respectively), with a 40µL/70µL ratio. Sorry, can't find the references immediately. Thomas Moritz used design of experiments.
John Halket showed that for MOX, you need 30 °C to ensure oximation, but don't want to go higher to avoid artefacts. Derivative yields increase to 37 °C, but Gln/Asn deamidate to form Glu/Asp above this (The work was done using model solutions and spiked samples).

[Consumer Products Guy]

I made/used trimethylsilyl derivatives for 40 years. We always assayed them same-day or overnight once autosamplers became commonplace. If we needed, non-derivatized extracts were reacted and assayed the next day.

I started using Tri-Sil in pyridine but I switched to BSTFA containing 1% TMCS by the late 1970s; this worked well for us so we kept with it. DMF or pyridine were used as solvents, with pyridine and 10 minutes on steam bath used when derivatizing active hydrogens on a nitrogen.

I even published several articles using BSTFA containing 1% TMCS.