When I worked in the organics preparation department in an EPA certified lab, we used hot soapy water (Alconox soap) rinse with hot water, then DI water, then methylene chloride then acetone. This removed all residue. Eash rinse three times. For headspace analysis, using commercially cleaned vials is cheapest way.
I agree with Karen's proposal to use new vials every time. You need to find a reliable supplier (we used Perkin Elmer years ago). Cleaning used vials is expensive in time and money, though in practice that may depend on how crazy your budgeting arrangements are (I've seen some).
A related question is how to dispose of used vials, unless they contain little else but water. That may be one reason for having to clean them anyway. I'm not aware of best current practice, but there is or was a similar problem with vials for liquid scintillation counting.
Note that some official pharmacopeial methods use DMF as diluent, which is classed reprotoxic under CMR rules. I don't thing heating this solvent under pressure in a glass vial should be allowed unless, perhaps, the instrument is under a hood. Some methods use DMI, a newer but widely-used solvent (inks and so on) that doesn't seem to have been as much investigated as DMF.
Thanks for your input Karen and Christopher. As I am attempting to identify the volatile compounds present in the headspace of insect infested plants, I am unable to use disposable headspace vials due to the size required (2L) so cleaning is my only option.
As far as your focus is on HS analyses, I strongly suggest to keep glassware (vials) and septa (I suggest the silicon/double Teflon layer type) in a dedicated oven at 110°C at all time (I am using a vacuum oven for additional isolation from the surrounding laboratory environment). Conventional cleaning of glassware as already indicated is definitively sufficient.
When I worked on the extraction and quantification of volatiles in wine at trace levels, I usually cleaned the glassware with RBS, rinsed with water, pure water and finally with ethanol (never with acetone which leaves volatile traces) then we put all the glassware in an oven at 400°C during the night. My colleagues who were working on volatiles of cheese (a food matrix which creates deposit on glass ware hardly difficult to take off) did the same procedure with success.
Andreas: vapour-phase silianisation is easier (especially for pasteur pipettes and so on), more effective and ecological. As described, a vacuum oven is needed, but I guess that was simpler at the time than using super-dry nitrogen or argon as carrier.
Fenimore, D. C., Davis, C. M., Whitford, J. H., & Harrington, C. A. (1976). Vapor phase silylation of laboratory glassware. Analytical Chemistry, 48(14), 2289–2290. doi:10.1021/ac50008a065
Best rinse solvent: this was discussed in books on vacuum technology. Apparently methanol leaves the least residue.
Thank you for your suggestion of silanizing the glassware before use. I just wanted to ask whether I would need to re-silanize my glassware between samples? i.e. after I have collected a sample and washed in detergent then water then dicholormethane then methanol?
I don't use deactived vials so I just wash in dishwasher and throw it in a muffle furnace at 550 deg C for 30 min to an hour. There will be no carbon compounds left after that guaranteed.