Hitting LOQs and Confirming Hits with GC-HRMS

Then: only two years ago...

I started working with the LASPB (Public Health Agency of Barcelona Laboratory) even before I joined officially two years ago. I got the opportunity because of my experience in high-resolution mass spectrometry – specifically Orbitrap-based MS systems; it was the focus of my PhD at CSIC (the Spanish National Research Council) and I’ve been using Orbitrap technology ever since. We are an official control lab and we analyze more than 35,000 (mainly food) samples per year at LASPB – so we’re pretty busy; in fact, we have an ‘open accreditation scope’, which means we’re obliged to analyze any food commodity that a client sends to us – including requests for new analytes. The lab here introduced LC-Orbitrap systems five years ago, which are particularly useful for confirmation or to troubleshoot problematic analyses, but GC-Orbitrap was unfortunately unavailable.

We were facing two main problems in GC-MS analysis. First, we found it challenging to reach very low limits of quantitation (LOQ) for some emerging compounds – polybrominated diphenyl ethers (PBDEs), a group of brominated flame retardants. In 2014, the European Commission requested that such compounds be monitored, with LOQ recommendations. Limited to a triple-quadrupole MS system, we had to work hard on sample preparation to concentrate the analytes of interest – if your instrument can’t catch the standard, you’ve got no hope in the matrix... We managed to hit the LOQs for all PBDEs except one – the notoriously tricky BDE-209. It’s a big molecule, which causes column difficulties but also sensitivity problems; sensitivity of triple quad instruments drops off significantly at higher molecular masses.

The second challenge in our GC-MS applications was in pesticide analysis – not because that is particularly challenging, but because, unlike LC-MS where we had Orbitrap-based systems, we didn’t have a confirmatory analytical method or an alternative technology for challenging matrices/interferences when it came to GC. Having confidence in our analytical results is extremely important, because they can have legal and financial implications; we must avoid false positives or false negatives!

Now: September 8, 2016

Now that we have the Thermo Scientific™ Q Exactive™ GC Orbitrap™ GC-MS/MS system, we still use the triple quadrupole instruments we did before for pesticide analysis – but if we get any doubtful or strange results, we’ve got advanced technology to dig deeper; GC-MS has finally caught up with LC-MS in that regard. The strategy with high-resolution MS is different to triple-quad methods, so we had to work hard on it. Analysis may be done in full scan, and certain parameters must be set. We developed a database that includes all the retention times, exact masses, and confirmatory ions – and from the database you can quickly set up a method for a confirmatory analysis. In fact, the work led to a poster, which received an award at the 2016 European Pesticide Residue Workshop held in Cyprus – a proud moment! I can also say, with some satisfaction, that I recently applied our protocol with great success when addressing an alert for propargite in oranges. We were pretty confident that most of the samples analyzed with the triple-quad method were negative for propargite, but one sample was not so clear. Was it actually positive? I quickly set up the method for our Exactive GC using our new database, and sure enough – we had a positive.

As for the PBDEs, we’re easily reaching all the LOQs now – including pesky BDE-209! The Exactive GC has good sensitivity and isn’t deterred by higher masses, and the high resolution allows us to pick out all of the interferences, so we’re obtaining beautifully clean chromatograms with very defined peaks. By creating a simple method with selected ion monitoring (SIM) windows, we can monitor not only the main peak but also the isotopic pattern – and we know for sure if we have a positive hit.

The upshot is that we now have the analytical confidence in the method to extend the analysis to other food commodities, beyond fish and seafood. On that note, I might add that we pushed the instrument really hard when it was installed – running 200 samples of salmon, tuna, and other fatty fish in the first couple of weeks. The instrument never missed a beat. We need such robust instruments.