Lowering injection volume doesn't drop area counts by same %

[pacerlaser]

I am using an Agilent 6890GC with a PDHID to analyze for oxygen, nitrogen, and carbon dioxide. I am doing manual injections using gas tight syringes. I have a gas standard with O2 at 10 ppm and N2 and CO2 at 40 ppm. I get very consistent area with replicate injections at 100 uL volume. However, when I do a 50 uL injection, the area counts drop only by about 15% from the 100 uL injection. I have tried different syringes (I have 3 identical gas tight 100 uL syringes) and all 3 result in roughly the same area. I thought maybe it was the column being a packed column so I switched to a capillary Stabilwax column and did some injections using a moisture standard. Low and behold, the exact issue remains; 100 uL injection results in x area and a 50 uL injection resulted in 0.85x. Does anybody know why an injection of half the volume would not result in roughly half the peak area?

[Peter Apps]

Depending what inlet you have, when you do an injection, the sample goes in multiple directions (split vent, column, septum purge) and in and out of some of the lines, and how much goes where depends on how much you inject and how fast.

Peter

[pacerlaser]

After doing more testing, it turns out that there is a background amount of nitrogen and oxygen while try to perform a manual injection. This is also true for moisture analysis using the PDHID. Looks like I'm doing to need some kind of Gas Sampling Valve to perform this analysis in the low ppm range.