GCMS newbie and the questions to be solved...

[nilo]

What ferrel type are you using for the column to the transfer line? It should be a vespel/graphite, full graphite will always leak. If they are new, once heated at the max temperature, they tend to shrink a bit and need to be tightened again.

>>Yup, i am using ferrul consisted by graphite and vespel. Yes!!! for the very 1st time it put inside the nut, after heated at high temperature (~300 degree celcius), air leak is detected when i performed tune at low temperature (~35 degree celcius), after tighten the nut agains, the air leak issue is solved and until now that GCMS intrument rarely facing the air leak problem.

I would not trust that library with that data.

Since you are heating before the run I would look more toward an injection problem (i.e. hexane is what you have in the clean-up solvents in your ASL turret?). How old is that needle? The liner? If this is extremely dirty you could cut an inch of so off the front of your column.

>> Hexane with HPLC grade, the needle is changed on every month basic (I had changed 3 new agilent gold standard needles). The liner also been changed every time. The agilent service engineer had cutted the column but still let the problem unsolved.

"i just wonder what is that (Siloxane from column ?)" The column will not give peaks, just bleed constantly and increase with the temperature.

>> Really ? then it should be weird because when i running the solvent (Hexane), i detected some siloxane compounds in between RT8-RT12. What do you mean by "bleed constantly ?"

The only thing that could suggest with the invisibility of C30 and up is that they are not reaching to the MS. Is the temperatures of your transfer line at 300C?

Normally when we run these types of runs we use a DB5 column and run it to 325C.

>> All problems have been solved, i'm using agilent HP5MS column


Good luck

Thanks for your reply and appreaciated your help.

[nilo]

Hi Peter, Thanks for your fast response ^^

Hi Nilo

I still cannot see chromatograms.

A very large peak that elutes early in the chromatogram may be due to the solvent - you can check this by looking at the spectrum. In GC-MS it is normal practice to turn off the MS filament until the solvent peak has eluted, and to turn it on before the first analyte peaks elute of course. You may need to extend the time for which the filament is turned off, or it could be that the solvent peak is either eluting later than it should, or is tailing more than it should. Late elution may be due to a carrier gas flow that is too slow, or a column temperature that is too low. Tailing of a solvent peak may be becuase the inlet liner or the top of the column are dirty, or because the split vent is opening too late (or not at all) in a split/splitless injection.

>> I had set the solvent delay to 4.5 minutes and during this period, GCMS filament is off. After 4.5 minutes, the Filament on MS side will lighting. I'm using splitless mode.

About the glassware cleaning - make sure that you rinse very well with clean water after you wash with detergent, and then after you bake at 500C do not rinse with other solvents, which can only add contaminants. If you leave an open beaker of solvent to evaporate down in a fume hood it will pick up all sorts of contaminants from the air. It is better to evaporate it under a stream of clean nitrogen.

>> Great Idea!!! I had tried so many times to clean my beaker with various cleaning method but still failed to get the nice blank. I will try to perform the solvent evaporation process under stream of N2 gas. I have to rinse the beaker with Hexane, as stated in the instruction before preparing the sample (my sample need to be rinsed with hexane and then the rinsage (hexane) is collected at cleaned beaker.

And you should not trust library matches unless you have supporting evidence for the identification.

>> Just wonder why the MS library cannot identify to simple hydrocarbon compounds. For examples,
It identifys a compound as tetratriacontane which have RT at 15. Tetratriacontane have 44 carbon (if i'm not wrong). When i saw the RT, i get suprised and checked the probabilty, OMG, the probabilty above 90%. I'm using ACCU Alkane standard and feel shocked

Peter

[Peter Apps]

The reason that an MS library search has trouble with simple hydrocarbons is that the spectra for all the linear alkanes are very similar - and the higher the MW the more similar they get. The molecular ion, which would give the MW, is always extremely small.

Peter