horible accuracy

[munio]

I want to determination of caffeine:
column HP1(cross - linked Methyl Silicone) nonpolar 25m 0,32 mm 1,05 film ticknes
flow 2,5 ml/min helium
temp inlet 250
temp det 300
gradient temp : 110 degrees C for 1 min
13 dg/min to 260
260 dg C for 1 min
5 dg C/min to 290
290 dgC for 5 min
Ihaven't autosampler, so - hand injection 0,5 microliter
I have bad reproducibility and Iwant to use a inert standard but I don;t know what I have to take , help me :)_ and sorry for my english

[WK]

Munio,
Can you give %RSD for injections? That would help - since manual injection will give bad RSD. But how bad is bad?
You could try using an internal standard if poor manual injection is suspected.
Lastly - have you access to HPLC? - since caffeine is not a great analysis by GC anyway.
Are you using a FID?
WK

[munio]

to WK
RSD is 13% so it's realy bad
and yes I use a fid,ofcourse I want to use inert standard but I don't know which of them will be the best

[WK]

Munio,
I would suggest that the poor %RSD is a combination of lack of internal standard and suitability of caffeine for GC.
Try using an internal standard and then you can see what the %RSD is.
An internal standard ideally should be similar in structure to caffeine - but I would just use anything to start with that doesn't elute with caffeine or sample peaks. Don't choose anything that is too volatile (like common solvents).
(You have tried using a new syringe and new septa?)
WK

[WK]

Munio,
Have you had any success with your problem?
WK