Increasing response from injection to injection on a GC/MS

[HPtarget]

I noticed sometimes on our 7890/5975 mass specs. The response for a compound would be around 50,000 and as the day goes on it will increase up to 300,000 and the next morning it will be back around 50,000. Also, we have internal standard and calibration checks and all compounds are doing the same. I believe the source is dirty since it has not been cleaned for about 2 months and we run about 300 runs a week on it. Also, the interface nut some times gets loose over time and helps, but I wonder is there a clear way to confirm with out venting the instrument. The 502 ion looks good also. I noticed also when we run clean samples response lowers and when we run dirty samples response will increase.

Any good information would be helpful.

Regards,

[James_Ball]

What type of samples are you analyzing?

If you are running samples with high levels of hydrocarbons like fuel samples, I have seen such problems before with those. Response increase with increasing analyte concentrations is a problem often asked about with the 5973/5975 instruments.

One article I found from Agilent says to use the EM voltage as low as possible to mitigate this problem, even using low EMV and lower split ratio when needed. Others suggest using the larger diameter draw out lens to help with linearity of the signals across the concentration range.

I have not used the new triple axis detector yet, but with the standard K&M horn I had this problem running volatiles and switching to the ETP film type detector seemed to help.

Is you instrument using a turbo or diffusion high vacuum pump? If diffusion pump, large changes in room temperature can cause changes in response. Warmer room gives less vacuum and lower responses, cooler room gives more vacuum and higher responses. I had problems with this before using a 5971 instrument and we only solved it with better control of the room temperature.

Oh for the interface nut, always use graphite/vespel ferrules and if possible I keep a few in the oven in a beaker so that the temperature cycling will preshrink them. Otherwise once you put one on and cycle the oven a few times you will have to re-tighten them a couple times as they shrink.

[HPtarget]

we have a 5975 triple axis.

I found we had a small leak in our press connector. And a small leak on the autosampler.
The interface nut constantly loosens. Thanks for the suggestion for the oven I will see if they can preshrink several ferrules.

However, we have air samples and only 10 ml aliquot of a max of 200ppbv standard with a mass range from 6ngs to 20ngs per a run. We have just received the larger Draw out plates.
I will ask the floor to try it and see if it works.

For the ETP, any recommendations? I will see if i can order one to see if that helps with some of our problematic instruments.

Tomorrow, I will the runs and see if they are still having response climb during the day. We run about 60 samples a day.

[Yama001]

Similar problems may also arise from adsorption or activity anywhere in the sample flow path. Standards and higher conc samples will mask it; clean stuff will lose analytes and give lower response.

[James_Ball]

we have a 5975 triple axis.

I found we had a small leak in our press connector. And a small leak on the autosampler.
The interface nut constantly loosens. Thanks for the suggestion for the oven I will see if they can preshrink several ferrules.

However, we have air samples and only 10 ml aliquot of a max of 200ppbv standard with a mass range from 6ngs to 20ngs per a run. We have just received the larger Draw out plates.
I will ask the floor to try it and see if it works.

For the ETP, any recommendations? I will see if i can order one to see if that helps with some of our problematic instruments.

Tomorrow, I will the runs and see if they are still having response climb during the day. We run about 60 samples a day.

Running air samples you shouldn't make the source dirty very quickly at all. I run about 40-50 drinking waters samples each day by purge and trap and I go at least 6 months sometimes 18 months between source cleanings and even then they are not very dirty. If you are getting air in the analyzer when it is at high temp though it will oxidize the surface of the source some, which usually turns it a bluish or gold color and that needs to be cleaned off with some abrasive.

You may want to start in the morning just scanning with the calibration gas valve closed in manual tune(closed so you don't have to wait for the cal gas to pump out before starting your runs) and measure the air and water counts at 18, 28 and 32, then once the response has risen to its highest level do it again and compare the results. If the air and water has dropped then it is a problem where over time the air leak is going away and you are getting a better vacuum causing the response to rise.

Is the oven temp low overnight when not acquiring samples or is it set at a mid ramp temperature? Cold oven could be letting the inlet fittings shrink enough to let air in at night then seal up once you begin cycling the oven the next day.