You are absolutely right in saying alkanes are volatile enough to be amenable to GC/FID or GC.MS analysis; besides alkanes have no polar functionalities and thus will not react with BSTFA. Plant epicuticle layers contain waxes as well as alkanes. So, like you I suspect papers reporting the use of BSTFA on solvent extracts of plant epicuticle layers do this to capture free long chain fatty alcohols and perhaps even free LCFAs.

Since BSTFA will not react with alkanes to form derivatives it will have no influence of on RTs of alkanes. For long chain fatty alcohols it will shorten RTs (decrease elution temperature) compared to RTs of free LCFAlcs. Derivatisation of polar compounds that could be run on a GC using high-temperature GC methods will also improve peak shape by reducing tailing.

You are absolutely right in saying alkanes are volatile enough to be amenable to GC/FID or GC.MS analysis; besides alkanes have no polar functionalities and thus will not react with BSTFA. Plant epicuticle layers contain waxes as well as alkanes. So, like you I suspect papers reporting the use of BSTFA on solvent extracts of plant epicuticle layers do this to capture free long chain fatty alcohols and perhaps even free LCFAs.

Since BSTFA will not react with alkanes to form derivatives it will have no influence of on RTs of alkanes. For long chain fatty alcohols it will shorten RTs (decrease elution temperature) compared to RTs of free LCFAlcs. Derivatisation of polar compounds that could be run on a GC using high-temperature GC methods will also improve peak shape by reducing tailing.

Thank you for good answer! Then, I need to identify unknown fatty acids, ketones, and alcohols by GC-MS with BSTFA derivatization. Also I need to identify alkanes without BSTFA derivatization in GC-MS. Then, I need to quantitate them with FID with standards. Is this right approach?

If you analyse your BSTFA derivatized extracts by GC/MS you should be able to identify all these compounds since their mass spectra will be different, in particular the molecular ion. You run your GC/MS in full scan mode (say m/z 40 to 400) and reprocess individual runs by having the MS software display the total ion chromatogram (TIC) in one pane and selected ion chromatograms in one or more separate panes. For example, TMS derivatives of long chain fatty alcohols yield a very strong (high intensitity) base peak at M+ - 15 (molecular ion minus a methyl) while in free long chain fatty alcohols typically undergo loss of water (18) and loss of of C2H4 (28) so fast, the first high mass ion one can see is at M+ - 46. The rest of the mass spectrum of a long chain fatty alcohol looks pretty much like that of an alkane (repeated loss of CH2 = minus 14). This is not the case for TMS derivatized long chain fatty alcohols. Apart from a strong M+ - 15 peak in the mass spectrum they also have a prominent peak at m/z 75 (dimethylSiOH+).

Example: eicosanol (C20H42O), FW = 298 g/mol; M+ - 46 = 252

Eicosanol-TMS (C23H51OSi), FW = 370 g/mol; M+ - 15 = 355

Since you have a fair idea about the compounds you expect to find in your extracts, you can use knowledge about their mass spectra to carry out specific searches of your TICs for presence of the compounds. While your TIC will show a lot peaks, the selected ion chromatograms (SICs) will show only a few. Staying with the above example, peaks in two SIC traces (one set to m/z 75; the other set to m/z 355) should only coincide once at the same retention time. The corresponding peak in the TIC (i.e. at the one retention time where a peak on the 75 trace coincides with a peak on the 355 trace) will be the peak of eicosanol-TMS.

Thank you very much again! In the FID, alkanes and derivatives of alcohols shows similar intensity or still dramatic different. When I see the GC-MS chromatograph, most major peaks were alcohols and alkenes as you mentioned. I was very difficult to see alkanes. If I use SIM mode in GC-MS, does it help to identify the alkanes?

Are there any tips for identification BSTFA derivatives and alkanes, which has low intensity?

EI mass spectra of n-alkanes all look fairly similar; the main difference is the molecular ion; however this is typically of low intensity. It is easier to find them in a complex chromatogram if you set up several selected ion monitoring (SIM) traces for fragments m/z 57, m/z 71 and m/z 85 you should be able to find (identify) the alkane peaks in the total io chromatogram.

BSTFA derivatization of polar functionalities such as OH-groups in alcohols create O-Si-(CH3)3 groups (trimethylsilyl or TMS for short). One product of their fragmentation in the ion source is the (CH2)2SiOH+ ion, m/z 75, which has a good intensity and easy to pick up by SIM.

For your application GC/MS is the method of choice since it enables you to identify individual peaks by their mass spectra and the mass spectra of alkanes and TMS derivatized LCFAlcs are sufficiently different.

Which fragment ions can I set for detecting aldehydes, ketones, fatty acids in SIM mode? Could you suggest fragment ions for SIM mode?