How do we get optimal chromatography?

Practical data from GCMS analysis.

In these years discussions in GC-analytical fora often relates to the use of Helium (He) or a transfer to Hydrogen (H₂) as a carrier for the GC.

We have analysed this a bit beyond the traditional pros and cons.

We made a mix of some of the common chlorinated pesticides used in the second half of the twentieth century. These were now analysed on one of our GC/MS systems. First with He and comparing this with H₂ as carrier using the original set-up of the ion source. Then changed the ion source as recommended by the supplier and comparing again.

We discuss here the chromatographic profile and the quality of the collected mass spectra. The machine used was an Agilent 8890/5977, a conventional quadrupole GC/MS with a relatively high-end (InertPlus/Extractor) ion source, with nothing changed to this.

Chromatographically we aimed for an optimum flow for the He system 35cm/sec. When changing to Hydrogen we had to change the inlet pressure to keep the linear gas rate down to around the 35cm/sec, and use pressure pulse during injection due to the lower inlet pressure, but except this: nothing.

We ran the components using He, changed to H2, waited two days to get the system “reasonably clean”, ran a series of pesticides, from Lindane to Endosulfan, and compared the chromatograms, see below.

• 1: Pesticides-Helium;
• 2: Pesticides-Hydrogen;
• 3: Alkanes-Hydrogen.

Fig 1

Table 1

The pesticides are written in elution order.

First results

As seen, Fig. 1, when using He the pesticides analysis is near-to-perfect (He: 1), whereas there is significant tailing on the hydrogen carrier chromatogram (H₂: 2). This has previously been attributed to the degradation of the analytes in the inlet, we lean more towards geometrical effects which inhibits the fast removal of large components from the ion source.

In (3) is seen a series of alkanes injected under the same conditions, and you will notice that below C16 (MW<=200) the peak shape is good, and above peaks begin to tail.

Another important parameter is the Score for the found components when searching in NIST or other commercial libraries.

We have used the routine NIST search (rev) in MassHunter, and in the columns below you see the Score values when using He or H₂ as carrier.

The score relates to 100%, and “ – “ indicates that the component is not between the 10 best hits. The results are seen in Table 1, the Score for the mass spectra when using Helium is “OK” without being ostentatious, whereas, when using Hydrogen, the mass spectra scores were relatively poor in most cases.

Further investigation

After this we went back to the lab, rebuilt the GC/MS system according to the recommendations from Agilent.

This means cleaning the ion source, and replacing the drawout plate (3mm) with a 9mm. Specifically, after the change to H₂ we waited six days before running the analysis to ensure that the ion source was reasonably clean.

Then, re-installing the 3mm again and run using He.

What did we learn?

The results (chromatography and data) are seen below (He: 1; H₂: 2), and as seen in Fig. 2, the chromatography is now looking good when using both He and H₂.

Also, the NIST search Scores, see Table 2, have seriously improved when using H₂.

However, the Score values for H₂ are consistently below the Score value when using He, to be more precise x̄(He) =73.2 and x̄(H2)=65.4.

Also in Table 2 we have now included the baseline Width, and we now find the Width on the same level for both He and H₂ (even though we did not optimize the H₂ flow rate), indicating that the “geometrical factor” is diminished by using the larger drawout plate.

Fig 2

Table 2

Conclusion

Overall, we think we can turn this into a series of recommendations:

• Our data indicate that the Agilent Inert Plus (Extractor) can be “rebuilt” to H₂ carrier with good results. “Good” means both chromatography-wise and mass spectra-wise.

• You should test using H2 on your type of GC/MS, it only takes a reduction valve and smaller H2 bottle. A qualified guess is that small molecules (<C12) will be ok on most systems.

• If specific recommendations exist for your MS regarding re-building, follow these.

• For “unknowns” analysis, for H₂, it should be considered whether the possible lower Score for the library search may be disqualifying for your type of analysis.

• For the Agilent 5977 version “stainless steel, turbo pump” and “InertPlus” a complete ion source which is  optimized for H2 exists. This ion source also fits the Agilent 7000 QQQ system. Otherwise you may rebuild the InertPlus ion source version with good results (as seen above).

Vi kan også hjælpe med