Thanks Mr. Christoph.The water trap was a new one,and ca.20 L-alanine standard only were proceeded.And I have check it's not yet deliquescence.Hence, maybe water trap was no problem; copper was also check and not yet completely oxdized, only ca.4 cm copper of all 34cm had already oxdized.

Further more, where is the possible source or determination of 29N2 ?

Before the needle valve of IRMS was open,the volt value of 28N2 and 29N2 were both zero mv.

And after the needle valve was open, the value of 28N2 and 29N2 were 11 and 38mv at the 1st time,and the latest time they were turn to 31 and 35mv respectively.

Which one above would refer to the real 29N2 BGD (background volt(mv)) ?

Is there anybody would help me to clarify it

What's the purity of your carrier gas (He)?  I can not remember clearly but it might be 99.999% or better.  Another possibility is from the Conflo, because the positions of all the capillaries are quite tricky.  You can try to slightly and gradually move the capillaries and check.  But be careful the moving must be very tiny (less than 1mm each time).

I agree with Christoph, it seems to be a problem with the water level. Which is your water level?. Which is the He flow you have set?.

Thanks ， Mr Jose. Do you means the background volt of 18H2O ? It's around/below 5000mv after stable. And Flow set is as fellow:

He-carrier: 300 ml/min     Oxygen:175ml/min    He-reference:100ml/min    

when analysis was proceed, the actual flow of He is stably 100 or 99 ml/min

Is there any mistake which effect the BGD of 29N2 significantly?

5000 mV for m/z=18 is too high. I should be below 2000 mV. You have a problem of water that it is affecting mass 29.

Dear Weiping Mei,

hope your case had been resolved. But if you have further technical problems with IRMS, you can go to http://isogeochem.wikispaces.com/ and qet friendly and professional support from colleagues. Start from searching archives that contain virtually all possible questions and answers.