new to GC/MS with a number of questions

[hplc.chem]

Dear all,
I am new to the GC/MS and its common practices. We have a Varian Saturn Ion Trap instrument, and I am trying to separate and identify some natural essential oils, in addition to some synthetic compounds.

For the essential oils, I run the sample then I print a report with the separation parameters for each peak and its related compound identification data based on the Fit value with NIST98 library. The problem I faced is that most of the FIT values are around 900 (out of 1000 value) or less.
My question in this part:
1. Is it common to get these Fit values? How can I improve the Fit?
2. Is it a common practice to rely on the Identification results produced by the library search?
3. What is the common acceptable threshold for the Fit value?

The other part of the story is identifting synthetic compounds, most of which are new ones. The synthesis collegues asked me to get the spectrum and the molecular weight. I ran the sample and separated the major peak from other tiny peaks. I then printed the spectrum of the major peak. However, In the EI mode, I could not find the molecular ion, so I tried the CI mode, because I read that it is a soft ionization technique. I was expecting to find the molecular ion as a major ion, but instead I got many fragments with the absence of the expected molecular ion.
My questions:
1. Is it common to find the molecular ion in GC/MS?
2. What is the best technique for obtaining the molecular weight of a synthesized compound?

Sorry for the long story, but this forum is really my best place to find the right expert answers for my chromatographic problems.

Thanks in advance.

[Hornet]

My question in this part:
1. Is it common to get these Fit values? How can I improve the Fit?
2. Is it a common practice to rely on the Identification results produced by the library search?
3. What is the common acceptable threshold for the Fit value?

Hello, i'll try to answer your questions:
1. Yes it is, you can improve your library search by making a full scan just around the interested range. For example if you are trying to identify light solvents there is no point in scanning masses above 200-250. Scanning high masses (especially when you reach high temperatures) means you'll get a lot of column bleed masses (207, 281 and so on) which will interfer with your library search because they'll be used as characteristic ions of that peak.
Another way is to do a background subtraction if you have too much interferences.
2. I don't really know if it's common to rely on library, in my opinion a compound should be checked comparing its retention time with a standard and then using the library for confirmation. Remember that library can't help you with isomers (PAHs, Xylenes, trimethylbenzenes are just examples) but retention time comparison + library can give you the correct answer.

[hplc.chem]

Another way is to do a background subtraction if you have too much interferences.

Thanks Hornet for your reply. Your answer brought a new question!
I found in the tutorials Manual the topic "background correction", but I don't know what the benefits of this procedure. Also, how can I do it in practice? Is it by injecting a blank and subtracting all the peaks appearing in the blank chromatogram?

[Amirav]

HPLC.Chem
About your new synthetic compounds, please note that about 30% of the NIST library compounds have either very small or no molecular ions and for synthetic compounds with molecular weight larger than m/z=300 only half of the compounds have reasonable molecular ions. CI provides molecular ions to many of those but not all and some compounds degrade on the closed CI ion source (close ion trap ion source).
We have a unique GC-MS system named 5975-SMB that provides molecular ions to practically all samples plus the EI fragments plus confirm library identification via isotope abundance analysis software plus can analyze twice bigger compounds than standard GC-MS
If you have a real challenge with your synthetic compounds that is very important for you I could analyze your sample (free).
For this please communicate with me over Email: amirav@tau.ac.il
and write some more sample (and your) details
I will also be happy to send you a paper on "identifying synthetic compound" and how to optimize reaction yields via our GC-MS
Best wishes
Amirav

[Alexandre]

Welcome to Forum and secret society of MS, young man.

Baseline correction can be done by clicking in the area of your peak then in the area of steady baseline next to peak, then click “Subtract baseline”. The software does the trick. You can do area subtraction also.

It may be different in your instrument, read specific help instruction.

[Don_Hilton]

You have entered a world where people with advanced degrees believe suddenly believe that the instrument that analyzes everthing on CSI actually exists.

There have been several discussions on using library search results on this forum, with some cautions on match score -which I presume correspond to your fit scores. Without attempting to repeat the discussions, there are cases where you can have very high match scores and a total miss -- and you can have a poor match score when you have the correct compund. (The later typically being the effect of noise on a weak spectrum)

For an understanding of what kinds of materials are likely to show a molecuar ion, and why others do not, I would suggest that you sit down with a book on mass spec interpretation.

There are refrences in this forum as well. I'd look up some links, but need to be leaving for work in a few. I'll try to check back this evening, when I get home.

[hplc.chem]

Many thanks Amirav, Alexandre and Don_Hilton for your help.

Amirav:
I appreciate the information you provided and your offers. However, we will not afford enough money to buy a new instrument currently, but I will keep this instrument in mind.

Alexandre:
The MS society is really a secret one, at least to me!
When performing "subtract baseline" what does the software subtract? does it affect the peak area or the MS spectrum of the peak? How it works i.e. What is the TRICK?

Don_Hilton:
I read some of the posts in this forum, but some are not clear enough to me due to my little practical experience in the MS world.
I appreciate your help with the links, and your advice regarding studying a book in MS, but do you recommend any book which is easy to follow and starts from the beginnings?
You mentioned the phrase "Weak spectrum", can you please clarify to me more? When do we get a "weak spectrum"? Does the concentration of the compound affects the spectrum?

Sorry for my many questions, and thanks for your experienced answers.

[MMJ88]

The baseline refers to the area between peaks, where there is no analyte eluting. You will still have a detector response though. In your software, click to view the mass spectrum at a few points in the baseline (i'm not familiar with the varian software, probably right click or double right click). You'll see there's quite a lot going on! When your analyte peak elutes, all the junk you see in the baseline remains. So the mass spectrum you're getting of your peak is representative of your analyte, but also a bunch of stuff you don't want. When you subtract the baseline, the software takes note of the mass spectrum of the baseline and subtracts that from the mass spectrum of your peak, (hopefully) eliminating the stuff you don't want and leaving you with a "cleaner" signal.

I believe what Don is referring to by a "weak spectrum" is a case where the signal (the spectrum of your analyte) to noise (spectrum of the baseline) is not very good. In this case, the matching software can give you very erroneous results. It takes the overall observed mass spectrum and tries to match it to a single compound. Perhaps someone can explain better than I can if you don't understand. As has been mentioned, there are quite a few discussions of the use of library matching on this forum already. I think you'll learn a lot from reading up on those.

[hplc.chem]

Thanks a lot MMJ88,

The background subtraction seems much clearer to me now. It seems to be an important step before making a library search, so is it "MUST" to do it for every identification? common practice? or just when we get a poor fit?

[Don_Hilton]

Background subtraction is very helpful, but if misued can lead to errors when trying to spearate signals from coeluting peaks - and sometimes you may not be aware of the coelution. I have joked that with careful placement of the points for taking the spectrum and background, one could create any spectrum he (or she) needs to prove a point.

I like programs that perform spectral deconvolution - AMDIS is available at no charge from the NIST web site. It is a bit tricky to learn to work - but can be well worth the time.

For books - firsts be sure you are comfortable with organic reaction mechanisims. Mass spectral interpretation involves the understanding of how molecules fragment - and requries an understanding of what can happen in a molecule. With an understanding of organic chemistry, then on to mass spec- the classic is McLafferty and Turecek Interpretation of Mass Spectra. (I've had my copy for years - and still drag it out now and again.) I've seen a text on mass spec coauthored by David Sparkman Introduction to Mass Spectrometery, browsing the table of contents - it covers the fieled well. Sparkman writes well, so if I needed an introductiory book on Mass Spec these days, I'd probably get a copy of that one.

[ods-at-pacific]

I have an idea that you may want to consider.

I have just started (today Monday January 16) a Web course on the interpretation of mass spectra. Among other topics, it covers how to use and how not to use the results of library searches. All the sessions are recorded and posted on the Internet for the exclusive use of the participants. The course comes with a 500+ page manual and other stuff, like a quick reference quide. You can go to http://www.mass-spec-training.comto sign up. If you are going to be doing this kind of work, you should really consider this course.

It does not matter that you missed the first session, you can watch the recording. The course is six week and meets twice a week. The next session in on Wednesday of this week.

[hplc.chem]

Many thanks to all of you for the replies. I think I will start intensive reading first, and I may then try to attend some training if I got enough support.

Many thanks again.