Building an MS database of fragrances

[csiquet]

Hello!
I'm a new member here on this forum even though I have been doing chromatography (HPLC mainly) for quite a while already. I hope to find answers to (some) of my questions.
Cheers!
--------------------

So, I recently started to work with fragrances extraction from plants, GC-FID/MS (Varian CP-3800, Saturn 2000) and fragrances identification for my PhD. I'm rather new at all of that...

I have found myself frustrated with the mass identification software as it gives me numerous possibilities for the peak I'm trying to identify.
I wish it would try to match retention indices as well... but it seems it does not have that feature. An alternative software with that possibility would be greatly appreciated. Ha! And change the x-axis from retention time to retention index.

Right now, I'm trying to built a mass database with the compounds we have available for the particular column and method we are using. I would like to know if there is a method/standard/protocol to follow to build such a database. For example, we have the FFNSC2 and the NIST databases. I suppose those mass spectra where taken with specified conditions for: sample concentration, EI parameters,...
I did notice that if a ask a spectral match on different heights (different concentrations) of one peak, I receive different possibilities. I suppose again that, due to the concentration of the compound, different fragments concentrations happen and the matching standard mass spectrum might be different.
This is why I would like to standardise the standards concentration and MS detector parameters in order to build the database.

Thanks a lot for your help!

[James_Ball]

I would imagine most spectral databases use spectra taken from pure compounds using a direct insertion probe so there is no background from a GC column to interfere with the pure spectra of the target analyte. Some are also generated from a theoretical process that produces what should happen in the mass spec to eliminate the variances in how the instrument is tuned.

You will never get a 100% match to a spectra from a database simply because there is no way to exactly duplicate the conditions from instrument to instrument. When you make your own library you will have the best match since it will take into account the background of the instrument and the specific tuning parameters you have set. Concentration will depend on the sensitivity of your individual instrument to each compound so it could vary greatly between each one how much you need to be able to see it well.

[Forse]

Your Saturn 2000 is a ion trap type MS and most spectras in commercial libarys are made with quadrupole type instruments, so your spectras will not match 100%.

[csiquet]

Thank you very much for your insights.

[Consumer Products Guy]

Our company had perfumers and a fragrance department, they had a few thousand pure compounds and also essential oils. That department felt it was important to identify fragrances in competitive products.

This was about 1990, and not much available commercially. We had an Agilent GC-MSD system. My co-worker and I injected about 20 pure components each night, then the next day cataloged them into a Search Library. Retention index was determined comparing to a mix of ethyl esters of fatty acids, because that's what the perfume industry used; also a DB-1 or DB-5 column was typically used. When available, structures and molecular weights were included. Minor components were included as well, and retention index data was needed as some fragrances had similar spectra. This search library was very useful constantly; the bulk of this library took us half a year to make, so somewhat costly, but was not like full time for that period!

Most of the ethyl ester standards could be purchased commercially; some we made from fatty acids.

In the US, most perfumers are clustered around New York City, and they use common names for lots of materials, so these common industry names were used as primary name in the search library.

[GOM]

Hi

I also have some experience in this area.

Before commenting further.

a) Unless you have a co-eluting peak you should get a good hit with the NIST database after baseline subtraction, whatever the peak height

b) What are the plant extracts that you are interested in? I have the ESO 2000 database of essential oils
see http://www.leffingwell.com/baciseso.htm

Kind regards

Ralph