Sensitivity issues with NCI

[SylvainL]

Good morning... here's a new set of problems

Here's what I'm running and testing

We're analyzing for BDE/BFRs using GC-MS in NCI mode (GC 6890, MS 5973)
column - DB-5HT with FS deact hi-temp guard column
const. flow @ 0.8
inlet is pulsed splitless @ 280C, with a dbl tapered galss liner
injection is 1uL
source is at 250C
quad @ 150C
and transfer line @ 300

our tune report is good but our sensitivity when running samples has gone down ~10x.

I've checked for air leaks and all seems well.

My question... sensitvity was definately better prior to installing guard column. I presume this is the issue... is there anything I could do or should I just revert back to no guard column?

our peaks tail a bit but they always have with bde/bfr.

Thanks

[jh1]

When you mention checking for air leaks is that with a leak detector or in PCI tune mode?

An Air leak, can be a really tiny leak but if the water PCI 19 m/z becomes too large our sensitivity really drops off.

[SylvainL]

Hi JH1, thanks for the reply. I've tested with a detector and in PCI mode. What value should I be getting when looking at 19?

thanks

[SylvainL]

just ran a manual tune in PCI, scanning for 17 and 19 m/z... wow 17 is over 1300000 abundance and 19 is at 10 000. the values are pretty stable around those numbers... I'm baking and flushing the column now hoping that it's just water in the column.

[jh1]

I normally try to keep it below .2%, typically I fiddle with any obvious weak leak points until I get the 19 peak down to about the size of the 18. However under .7% isn't terrible especially for an older instrument, I've noticed as instruments age the tiny air leaks you can never track down slowly multiply.

I suppose you can always pull out the guard column and see if your % water drops significantly.

[SylvainL]

after everything stabilized... I was at 2% ratio 19 to 17m/z. this would be indicative of a leak I assume?

thanks again jh1

[jh1]

after everything stabilized... I was at 2% ratio 19 to 17m/z. this would be indicative of a leak I assume?

thanks again jh1

Yeah that's a bit high, I'd try removing the guard column and see if your ratio drops significantly.

[SylvainL]

I'm really close to scraping the guard column... the only prob when I go back to when the prob started, the guard column had been on for over 3 months. I've tried to pin point if anything happened between good sensitivity and deteriorating sensitivity. It just randomly changed... mind you I'm going on the log book so maybe someone did something without logging it in.

quick q for you... if my tune report seem good, this would indicate a prob on the gc side would it not?

thanks

[jh1]

I'm really close to scraping the guard column... the only prob when I go back to when the prob started, the guard column had been on for over 3 months. I've tried to pin point if anything happened between good sensitivity and deteriorating sensitivity. It just randomly changed... mind you I'm going on the log book so maybe someone did something without logging it in.

quick q for you... if my tune report seem good, this would indicate a prob on the gc side would it not?

thanks

Sometimes the connector between the guard column and column loosens enough to allow air in just due to expansion and cooling between sample runs.

We've actually been using DB5MS+DG columns, columns with guard columns fused on for our CI in order to cut down on air leaks since it really hates water, however our method maxes out at 320C and Agilent doesn't make your column with a fused guard column unless you have them custom make it.

Typically if your tune turns out okay the MS is normally working fine, if you want to double check if your source is dirty and dirt is causing your signal loss you can ramp the senses and see if any of the ramps look especially choppy, this can sometimes help you identify if the source needs cleaning.

[SylvainL]

Hey JH, when i read your post this morning I figured I might as well take a look. I had checked the connector and it seemed fine but figured I'd redo the connection. What I hadn't noticed was the warning on the connector doc about not not exceeding 325C or degredation will occur. Our method ramps up to 325 and stays there for 7mn. With the amount of samples we run I wouldn't be surprised is this is the cause of my gray hair. Anyways I've contacted restek pertaining to this and will keep you posted.

Thanks for everything

1.5 days to summer holidays

[SylvainL]

update... I removed the press-tight connector, thereofre removing the guard column completely. Reconnected, ran ncich4 tune (all good)... ran a std and the sensitivity was still at the same value.

The only other area would be to recheck the end of my transfer line maybe the column tip was damaged when I closed the detector.

so far I've done the following:
cleaned source... al oxide rub and sonicate
changed transfer tip seal, repeller insul
reinstalled column going into transfer line... new ferrule and no more than 1mm pass tranfer line
changed inlet septum, glass liner, liner ring, gold seal and washer
removed guard column and reinstalled column with new ferrule

It almost seems like something is slowly deteriorating.

Tune report always good.

[jh1]

Hmm, so it's not a leak at the recently installed guard column.

Could you post a copy of the NCI, PCI tune reports and a picture of the 17,19 m/z peaks in PCI mode? It just seems like you're having an air leak coming from somewhere, those are very classic signal killers with CI and your water peak seemed high from your description.

A couple other obvious things that you've probably already ruled out:

- Make sure the nut into the auxiliary is nice and snug, that little sucker always likes to loosen up on me and let a tiny bit of air in.

- Make sure the vapor traps on the methane and carrier gas lines aren't leaking and have been changed in the last couple years, these do go bad after a long enough period of time and can then let water in.

- Check over your method and make sure someone didn't change something they shouldn't have source temp, split ratio etc.

And when you cleaned the source did you make sure the tiny hole below the filament in the source body was clear? That thing loves to collect dirt and sometimes I have to run a wire through to push out all the crud because just cleaning and sonicating doesn't always do the trick, although normally this results in decreased area counts for your tune ions as well.

[SylvainL]

ok new info... I've been running a standard and the sensitivity gradually decreases with every injection...

jh1 thanks for the reply again... the nut to interface is nice and snug and I've checked for leaks from the cylinders to our GC (all good). I'm venting right now (myself and the MSD), i will clean the source again (I wasn't the one who did it last so it could probably be that).

Btw the column must extend approx 1mm past the transfer tip right? and not past the transfer tip seal tip? (that's in CI mode)

Also we've ran about 400+ samples on the existing column... so if all else fails I might try a new one since we do have one in stock.

Thanks again and I will keep you posted

also, sorry for the delay in responding i was on holidays last week... wish I was still there

cheers

[aldehyde]

You should check your tuning mass abundances and compare after doing a couple runs, do a few more runs, compare, etc. If your analyte sensitivity is dropping but the tuning abundance stays the same it may be due to activity in the inlet or column.

Dealing with the same kind of problem myself .

[SylvainL]

Ok here's yet another update... still not great.

I cleaned the source and retrimmed the column entering the detector. Ran leak test in PCI mode all good no signs of air/h20. Ran a CI tune, all good. Injected my standard and got phenomenal results... sensitivity had shot back up (~10X). 2nd injection(same standard and only about 30mn apart), BAM sensitivity down 1/2, 3rd injection down a little more.

Aldehyde sorry to hear that your having the same prob... it's frustrating that's for sure. The inlet (liner, seals, ferrule and all) has just been changed and or cleaned, it was part of this whole trouble shooting.

What i don't get is why it went down to 1/2 after one run?? I retightened the nut at the interface in case of leak but still same prob. Could this be a reagent gas issue? I have a hammer in my truck maybe that would fix the prob... if it doesn't I'm sure I would feel a whole lot better.