[av0930]

I have a couple of questions for all you Mass Spectrometry experts and would greatly appreciate your input:

1. Should the retention time (RT) on a spectra of a specific substance (benzoylecgonine for example) be the same, or similar, no matter which laboratory or which GC/MS machine is being used?

2. Should the spectra look the same, or similar, if it is the same substance?

3. If the retention time on two different spectra of the same substance are similar, and a third spectra of the same substance has a different RT (by 4 seconds), can it be concluded that the spectra with the different RT is not actually the same substance? By the way, the two spectra that are very similar are from a hair sample and a urine sample. The third was a urine sample. All three tests were done at different laboratories.

Thanks in advance.

[lemonoman]

If the exact same amount of the exact same substance was put into GC-MS's under the exact same conditions, and the GC-MS's were calibrated exactly the same...then the exact same spectra should be produced.

Unfortunately, this isn't the case in real life

I suppose they SHOULD look the same, or similar, and have similar retention times...but if the temperature isn't calibrated right on one of the machines, or if there's a piece missing from a GC column, or something else, it probably won't be exactly the same.  In fact, it only takes a single difference in the conditions to make the two RTs differ.

And so, just because two substances don't have the same RT - especially if it was done at a different laboratory - you can't definitely conclude that it wasn't the same substance.  You'd have to both run a separation/MS on a sample of the exact same substance (i.e. stock hexane or a mix of alkanes) and make sure that both machines give the same reading for that standard.

Hope that helps

[av0930]

Thanks for the info lemonoman.  If I posted the GC/MS information would you be able to tell why the retention times were so different (difference of 4 seconds)?

[lemonoman]

As humbled as I am to say it, I couldn't.  I haven't worked with it enough to be able to do that

[Dude]

1.  The retention time of a sample will be dependent on the GC conditions.  The length of the column, the carrier gas flow rates, the stationary phase thickness, the temperature ramping profile, etc will affect WHEN your sample comes out.  The sample retention time may vary by an hour and still be a valid measurement with good resolution if one lab uses a 100 m column and another lab uses a 10 m column.  Before making any conclusions, gather this information for each lab and post it.

If both labs use exactly the same conditions (i.e. as specified by an ISO or ASTM method), the retention times should be similar (not exact but within 10 % of each other).

2.   If you are referring the the mass spectra, and assuming you are using electron impact ionization source at 70 kV, then yes.  The spectra should look nearly identical with minor variations for a particular compound.  Peak matching software is commonly used to associate compilations of intensity and m/z ratio to known compounds because the variations become very subtle when trying to match similar compounds (i.e. 2-methylhexane and 3-methylhexane).   

[av0930]

Dude:

Thank you so much for your input.  Now, you say that a Mass Spectra should be the same or similar assuming you are using electron impact ionization source at 70 kV.  Besides the 70 kV, is this still dependent on the conditions equipment to equipment? I found a mass spectra of Benzoylecgonine at this site http://www.varianinc.com/image/vimage/docs/products/chrom/apps/gcms67.pdf and it looks nothing like what is in the results.  The base peaks are different.  I am going to scan the mass spectra in the results along with all the gc/ms information I have and post it with the hope that you can assist me in determining why the mass spectra is different.  I greatly appreciate you input "Dude".  Thanks again.

[Dude]

As Lemonoman indicated, if the mass spectrometer is calibrated and working correctly (i.e. no air leaks, etc), then I have not seen one instance where a compound varied "significantly" from the NIST database of the spectra.  If the spectra does vary considerably and the mass spectrometer (they usually have some internal organofluorine solution if it is an HP or Agilent) is working correctly, then a large variation (especially in m/z locations) between the spectra would be a big "red flag".  You might need to do some other work or repeat the experiment and try to isolate as many variables as possible (i.e. run all three samples on the same GC-MS).  If you run a sample of toluene (for example) at 10 am, run your three unknowns, and then run toluene again and the toluene spectra are consistent and the toluene retention times are consistent, then you can infer that the GC-MS is working correctly.  If the same variation occurs with the three unknown samples, you can almost unambiguously infer that the compounds are not similar.  Most labs use this approach of some variant of SQC (statistical quality control).  That's the ideal case.  I realize you may have limitations in sample availability and other constraints so that's when it gets trickier.

[av0930]

Once again Dude, I thank you for your help and prompt response.  There are other questions I have (if you don't mind), but I need to post the analytical results so you can refer to it when answering my questions.  Although, one question I can ask is concerning the m/z peaks.  Do they also refer to them as ions?  For example, is Tgt m/z 318 the same as ion 318?  I'll post the analytical results by Friday.  I need to scan them so please bear with me.