poor calibration curve

Hi,

One of the test for the GCMS is VOC, and the method has 14 different VOC components, with 2 I.S.. The split ratio is 10:1. The problem I am facing is some components cannot meet the R^2 requirement, R^2 > 0.995. Sometime meets the requirements but sometime don't. I am not sure what causing that problem.

For GCMS, the tune report is fine, abound for m/z 69 is around 400K, EM volt is around 2300. regular maintenance of septum, liner.

For sample prep, 14 different individual std, analytical standard grade. usually make a stock (1000ppm, all in one flask) and then 5 different cal point, 2.5, 5, 10, 15, 20ppm. tried to share the volumetric flask as much as I can each time.

Talked to our engineer, from his point of view, if the tune report is fine, the gcms is fine. It gets me thinking how I can ensure the calibration curve pass the requirements (R^2 > 0.995) everytime.

0 paulsalverda 1 month ago

EM at 2300V is very high....how old is the E.M.? How long since the ion source was cleaned? How long since new filaments and insulators were installed?

Are you doing Purge and Trap or liquid injections? 10:1 split ratio with typical GCMS column flow at 1.2 ml/min is down there in the very low total flow range. The GC inlet works better at >= 20 ml/min of total flow. It might work down to 10 or 15, but the gas flow dynamics down at those low flows can cause issues.

In Windows Explorer there is a file underneath your method's VOC_05Aug2024.M (for example) subdirectory called acqmeth.txt. Please copy/paste that into a reply along with the hardware models and the data acquisition software version you are using.

If the system completes tuning and some compounds work and others don't --- it is not a detector problem.
Cancel
Vote Up 0 Vote Down

Sign in to reply

Verify Answer

Cancel

Was this helpful?