1. Fiber selection: Normally the DVB / CAR / PDMS 50/30 μm and PDMS / DVB, 65 μm fibers give higher peaks from the head-space analysis of plant material. But the 100 µm PDMS standard fiber is more robust. With this fiber you are able to proceed a high number of analyses runs.

2. The first peak after fiber injection is air and CO2 from the needle.

3. Sample preparation: It depends on your material and aim. Normally it is advantageous to prepare a defined sample by weight in a portion of plant material to the headspace vial. Additional it is profitable to use salt (NaCl) to saturate the sample (salting out effect) for higher peaks.

Additionally please find also SPME protocols in these papers:

1) DOI: 10.1021/acs.jafc.5b00704; ULRICH et al (2015)J. Agric. Food Chem. 2015, 63, 3348−3356

2) DOI:10.1021/jf1034948; OLBRICHT et al J. Agric. Food Chem. 2011, 59, 944–952

TO ME, I PREFER DOING THE SPME VOLATILE COLLECTION DIRECTLY FROM THE PLANT MATRIAL VESSELL,WITHOUT MOVING PLANT MATRIALS, FOR ANY PHYSICAL TOUCHES OF THE MATERIALS MAY RESULT IN ADDITIONAL CHANGES IN VOLATILE PROFILE. THE SPME METHOD IS GOOD FOR QUALITATIVE ANALYSIS OF VOLATILES, BUT NOT ACCEPTED FOR QUANTITATIVE ANALYSIS, UNLESS CALIBRATION  CUVERS ARE ESTABOLISHED USING AUTHENTIC STANDARDS. IN ADDITION, PLANT MATERIALS SHOULD BE CAREFULLY WEIGHTED TO HIGH ACCURACY SO THAT THE PLANT EMITTED VOLATILES COULD BE COMPARED BASED ON PLANT MASS.

a good source on SPME is Supelco who makes the fiber. Their technical support should be able to help you greatly, Please contact them.

Thank you very much for the suggestions

You said "... in the first minutes of GC-MS analysis, a peak (probably corresponding to CO2) appeared that disconnected the detector wire, then we thought to solve this by turn off the detector for the first two minutes."  I assume you mean the heated filament in the source. Although you will introduce air, one SPME injection will not cause the filament failure, it was just a coincidence and it was time to change it anyway.  Turning off the detector until the air has eluted is a good strategy to protect it.  Also keep in mind that SPME fiber injection is more damaging for the inlet septum than smaller needles, so you will need to change the septum more frequently to avoid leaks.  (Unless you are using a septumless injector.)

Dr Veazey, 

We had this problem by exposing the fiber directly to the culture flask, which we think is the detector's turn-off; As soon as the signal of this compound appears (it seems that saturates the column) it is not possible to observe more peaks, the detector's turn-off. After that, we transferred a small amount of in vitro plant to sealed glass vial for VOCs headspace microextraction and the analysis in CG-MS was done without problems. But I had not thought about the need to change the septum more frequently to avoid leaks.

Thank you very much for the suggestions.