Agilent GC/MS autotune problem

[ckeller]

We are consistently seeing peak(s) around 14-18 amu over 40 abundance in the autotune spectra on an Agilent 5973 MSD connected to an Agilent 6890 GC. We had a technician come out and the only solution to the problem is to do a startup.m method that starts at 60C, Initial Time =0, Ramp of 30 to 250 for a final time of 10 mins, run time is 16.33 mins. The Front Inlet parameters are Split mode at 250C, pressure at 5.30 psi, split ratio – 23.5:1,
split flow – 35.7 ml/min, total flow 39.3 ml/min. The tech checked for leaks through out the entire system

The general.m method that we use for the analysis is a bit different being in a splitless mode, start temp of 60C, Initial time 2 mins, rate 30C, 300C for 5 mins.. Front Inlet splitless 250C, 5.30 pressure, purge flow 35.7, total flow 39.3.

We created a new method called gensplit.m where the only thing we changed in the method was the split mode and the peaks were greatly reduced.

The questions, other than the problem stated above, are:
1. Why are we seeing these peaks? What could they be – water, nitrogen, oxygen??
2. Is it really a problem to run an autotune using a different method than what you are running the samples on? I don't think so as the autotune is checking the MS of the instrument. When I compared the startup.m to the general.m the only differences are in the GC parameters and none in the MS parameters.
3. Any advantages or disadvantages to running split or splitless?

Some people think that is moisture built up in the GC system and when we do the startup.m, the moisture is burned off as the peaks are lower if the autotune is done immediately after the startup.m methods ends. But could it be another contaminant?

The analysis on this GC/MS is for controlled substances, qualitative analysis only.

Thank you for your help.

Christy

[Peter Apps]

Hi Christy

These are classic symptoms of a leak in the inlet - when you go to split the extra gas flow thorugh the inlet dilutes the air that is getting in. The most likely spot for a leak is the septum - when did you last change it ?, next would be the column connection, then the gold seal. For the technician to have "solved" the problem by changing the analytical conditions does not point to high level skills on his (her ?) part. How did he check for leaks ?

Peter

[ckeller]

Thank you for your answer.

Septum change - that's the first thing we did. Didn't seem to help over time. If we change the septum, do a few injections, the leak is back the next time we use the GC/MS.

Stupid question - what and where is the gold seal?

The tech checked the connection, I believe, with the helium leak detector 'machine'.

Christy

[Schmitty]

If it is returning so quickly, it might be an incorrect installation of the septa, or a worn needle on the autosampler (I am assuming you are using a 7673 or something similar). Do the septa look punched out or torn up on the bottom?

The septa nut only needs to be finger tight.

[Peter Apps]

Hi Christy

The "gold seal" is a metal disc that is sealed against the bottom of the inlet body by a nut at the bottom of the inlet. You can see the nut (or its insulated covering) in the top of the oven. Changing the seal is a part of routine inlet maintenance - check out the users manual.

You have already confirmed that the septum is leaking by the improvement when you install a new one, so I owuld leave the seal for now and go with Schmitty's advice to prevent the septa being damaged so quickly.

Peter

[AICMM]

Ckeller,

Several comments. Do you mean to say that over 40 is over 40% in the autotune? If so, you have a significant leak. Second, don't overtighten the septa as this will actually cause it to leak faster. Set your head pressure and when you tighten down your septa, when you start to get pressure, a quarter turn more. Thats all that is needed. Third, moisture would account for the 18 but what about the other peaks. Sort of shoots the moisture build up theory. Fourth, the thing about the gold seal is, it is really the washer that goes underneath that makes the seal. Make sure you have that washer and make sure it is underneath.

Finally, regarding your tune. If you use the autotune to verify maximum MS performance and don't use it, why did you use it in the first place. If you don't use autotune for analysis (and lots of people don't in the environmental business) then what do you use to verify MS performance? This sounds a bit contradicotry but a lot of people in the environmental business use autotune as a starting point, the alter the tune so that they can pass EPA criteria. But the autotune is still the starting point and then there is a subsequent check. When you say qualitative, you mean just for confirmation but are you using ion ratios?

Regarding split and splitless, in your case, sensitivity is the major factor.

Best regards.

[ckeller]

Ckeller,
Several comments. Do you mean to say that over 40 is over 40% in the autotune?

Yes 40%

If so, you have a significant leak.

That's what we thought

Second, don't overtighten the septa as this will actually cause it to leak faster. Set your head pressure and when you tighten down your septa, when you start to get pressure, a quarter turn more. Thats all that is needed.

We'll try that out.

Finally, regarding your tune. If you use the autotune to verify maximum MS performance and don't use it, why did you use it in the first place. If you don't use autotune for analysis (and lots of people don't in the environmental business) then what do you use to verify MS performance?

The person who set up the daily check for the GC/MS has an autotune as part of the process. We are to look at various parameters in the autotune to see if the instrument is working correctly.

When you say qualitative, you mean just for confirmation but are you using ion ratios?

We don't do quanitation of drug, we are identifying the drugs by their retention time and the mass spectra when compared to a spectra in the library.

So would it make more sense for use to run a standard mix of alkanes or some other types of compounds to check on if the instrument is working correctly or not. I'm not sure what the autotune is telling us other than the leak. It's been a while since I used the GC/MS.
Thanks for your help.

[chemstation]

Hi,
Question,
1) Has the peak suddenly appeared, or has it been gradually increasing over time.(use the autotune printouts to check) ?

2) Is the gc/ms a new setup, or been moved/altered eg gas lines changed.?

3) to check the quality of the carrier gas, make methods with the following parameters:

Method A:
constant flow at 1 ml/min
gas saver mode off
splitless injection

Method B:
constant flow at 1 ml/min
gas saver mode off
split injection
split ratio 1:500

to monitor the peak abundance, select in the menu select 'Instrument','Edit Tune Parameters', & select [SCAN]

what should happen is that for method A, you will see no change.
In method B, if the peak goes away after about 2 minutes,
then the carrier gas is good and the 'peak' is entering into the carrier gas in the lines or injection port.

(N.B. you can select the Split mode and ratio on the GCs Keypad, saves having to load the tune scan again)