GC-TOF-MS detector voltage

[badrul]

Hello,

Just for clarification, is it true that you can double the signal every 50 V of the TOF voltage? Is it ok to increase the voltage to 2000 to obtain higher sensitivity? Is there any drawback to it? Furthermore, does the ion source and transfer line temperature play any part to the TOF sensitivity?

Thank you and really look forward for your guys comment on this matter.

[Don_Hilton]

From yoru question, I am guessign that you are using the LECO TOF, so I will answer in terms of the LECO TOF. While the signal does approximately double with each 50 V, the background noise grows also. And, it does not grow with the same function. The important characteristic to get the best sensitivity is signal to noise on analytes of interest. At 2000 volts, the noise may be sufficiently high that your signal to noise has begun to decline. And at 2000 volts, you age the detector more quickly. In the Pegasus instruments that I have used, I have found that there seems to be a sweet spot about 1800 V. And, with a new detector I do not need to go even that high.

The best thing I can suggest is that you check the sensitifity of your instrument by injecting 2 pg HCB and measuring the signal to noise. The specifiction on the Pegasus instruments in 1D operation is a signal to noise of 10:1 on m/z 284 when the peak width is one secon wide. Understand that a number of factors affect this value, starting at the inlet. This becomes a full system check (and is one that I do every day when I am running samples). You can then measure the signal to noise at several different detector voltages to see what the sensitivity is. You may discover that with a new detector you can get noticably better than this sensivity. While you can run the instrument at this better sensitivity, understand that you will not have this sensitivity for the full life of the detector and that the next detector may not be this good. Set your expectation at meeting specification and use the instrument at that level of sensitivity and you will get a good life out of the detector.

Transfer line temperature affects the sensitivity only in that if the transfer line is too cold, peaks may be broadened -- and broad peaks have less height (note for a given mass of material, aways the same area - so wider has to be shorter), or less signal relative to the background noise. Some of the earlier Pegasus instruments did not have good thermal isolation between the transfer line and the ion source. If you can run a transfer line at 270 and an ion source at 220, you do not have this problem.

Source temperature. In the past, I have tended to run an ion source at about 200, but have been told by some who have checked it out, that it gives better results at 220. Remember that increased temperatures increase fragmentation, so this may depend on which ions you are looking at in a spectrum.

For sensitivity - go for narrow peaks. Watch your linear velocity and run close to the optimum flow rate or just a bit faster to keep peaks as sharp as you can. for 1D chromatography, use a ramp that is fast as you can use while still obtaining necesary separation between peaks. Keep the inlet clean and all fittings tight. Watch the leak checks on the instrument. Run the leak check report daily and be sure that the intensity of m/z 28 is below 5% of that of m/z 69. I like to see this level at 2 - 3%. (You may need to be sure the GC is in split mode wiht at least a 35:1 split ratio to avoid intrusion of air through the split vent if you are using a column that requires very little pressure above atmospheric to drive the column.) I have heard the argument that the 5% level is too difficult to reach, but I have done this with several Pegasus instruments - even ones with pressfits installed as part of the GCxGC configuraiton. It just takes careful attention.

And, use the most current version of software available for the instrument. If you have an older instrument running a version before ChromaTOF 3.xx, upgrade it. And if you do not have the latest update of ChromaTOF installed, install it.

So, those are some ideas to get you started.

[badrul]

Thanks for the detail recommendations. It was very helpful. One more question if I may ask, is it okay to inject 200 ppb HCB instead of 2 ppb in order to check the instrument sensitivity? This is because some of LECO's application notes mentioned that sensitivity of the instrument was checked by injecting 200 ppb of HCB with a S/N of > 1000. Should it be okay to make this kind of assumption and comparison in order to check the instrument sensitivity?

[Don_Hilton]

Do it. 2 pg is the limit of detection. You have 4 or 5 orders of magnitude dynamic range on that instrument - so if you shoot more than 200,000 pg HCB on column, you should suffer nothing worse than a flattened peak. Shooting 200 pg is helpful in that if the column is getting a bit old, you have less risk of the HCB peak being lost under a huge silane peak.

[badrul]

Thanks for the reply. Really appreciate it. From your experience, how long is the lifetime of the detector if on average 1800 V is used as the detector voltage. Just wondering..

[Don_Hilton]

My instruments have been used intermitantly and I've had at least four years out of detectors. If you run the instrument 24/7 I could only guess - but no less than a year.

To make the detector last as long as possible, do not acquire through solvent peaks unless you have to and do not acquire low masses unless you have reason to keep the data. (The instrument will deflect masses 28 and 32 away from the detector if you do not acquire below some mass, which I do not remember. But if you may need to look at the runs at a later date, those lower masses may be important in retrospective data analysis, so think carefully abou the mass range for acquisition. Loss of important sample data is far more expensive than a few hours or days of detector life.) And keep the instrument leak tight.

[badrul]

Sorry if this has been posted before, but I've encountered some sensitivity issues when analyzing pesticide residues at low levels when using the 1D mode. Since I rarely try the 2D mode (I don't know why, maybe because of the liquid N2,or colleagues who always disturb me with management work), theoretically does the 2D mode gives significant increase of sensitivity apart of better separation of the chromatogram?

Hopefully next week I'll get the time to do some 2D work.

[Don_Hilton]

You do get better separation - which can help - and the column bleed from the first dimension column is focused and disappears into the solvent band region of the chromatogram - which gets rid of some bakground.

Also, as the 2D modulaton focuses the compund at the head of the second dimenion column, the peak is sharpened. If you take a mass of material that gives a 2 second wide peak and you focus so that it now gives a 0.1 second wide peak, you would expect the peak to be 20 times as tall (the area stays the same). Given real peak widths in the first and second dimensions, slicing a peak into multiple pieces in the modulator, the necessary increase in acquisition rate to acquire 0.1 second wide peaks you can expect an increase of 10 to 30 times in signal to noise for those small peaks that you have had trouble finding.

The tradeoff for 2D: If it is just sensitivity, I would suggest that you look at the sample prep and get more sample on the column. If you are looking for one or two compunds, consider good sample prep and good chromatography. However, if it is also a matter of coeluting compounds (and if you have multiple target analytes at trace levels in a complex matrix, coelutions become a problem very quickly, even with deconvolution software), the 2D chromatography is well worth the time and effort. There was a nice paper that came out a while ago showing the application of GCxGC-TOFMS to analysis of grapes and wine for pesticides and other environmental contaminants - and the list of target analytes was amazingly long and the sensitivity seemed to be good.

And if you want to do non-target screening, the GCxGC is phenominal - because you have to examine everythign and with good sensitivity. I had fun with a sample of houehold dust. We got an extract from the dust and ran it in the GCxGC-TOF, giving rise to over 10,000 peaks in the peak table. We then used the "scripting" feature in ChromaTOF to pick out anything that the software could find that looked like it might contain chlorine or bromine. We found all kinds of stuff. I can't give sensitivity numbers for the compunds, because some of the spectra are for things we never were able to identify with a name. I don't remember the particulars, but the details are published in J. Chrom. A as is the method for contaminants in grapes and wine (different issues of the journal, though).

And my digression is more to point out that there is a lot of fun to be had in doing serious work with GCxGC. (I have fun when I get good results, particularly when the results are "impossible" for other folks.)

[mg1773]

I have a quick question.

I have Leco Pegasus 4HT. Recently, I could not get any signal wen I tried to profile. Any idea what could be the problem. The push pulser is ok and the filament is brand new