Problems with Shimadzu GCMS

[LWC]

Hello,

This is in regards to a previous post, but the issue still has not been resolved if anyone has any other idea pls let me know.

Shimadzu GC17A MS QP5000

Below is a little history:

Shimadzu PM 3/3/09

Terra Tech PT PM 3/11/09



In the 2nd week of May we started to lose sensitivity, the ISTD area counts went from 1,000,000 down to 200,000. I made new solutions/recalibrated and this did not improve the results. Tune file showed a slightly dirty system, peak 502 jagged. I started tech support with Shimadzu:

6/8 methanol rinses and system bake out, no improvement
6/9 complete leak check, small leak found, tightened fitting, no improvement.
6/9 direct injection results not any better than PT injections
6/10 changed He tank, no improvement
6/15 cleaned ion box, slight improvement
6/22 changed out inlet parts/consumables, no improvement
6/30 replaced the column, saw immediate improvement, but only lasted a few days and returned to initial state
7/6 contacted EST made some pressure adjustments, saw improvement, but again didn’t last.
7/10-7/19 on vacation
7/21 cleaned lens stack, noticed it was installed backwards, from PM in March corrected issue, saw immediate tune improvements, but sample runs have not improved
7/24 installed new traps in PT, no improvement

I have done direct injections, no improvment. I also recently had a Shimadzu Tech in here and he installed a different column and it passed specs. We reinstalled columns that I had and after multiple runs it returned to a lower sensitivity. We have not tried multiple runs on a different type of column due to time constraints. Another contract lab said our quads needed to be re dipped and the analyzer replaced, however the tunes look fine. The tech is going to return next week to look at the isssue again if anayone has any ideas I could pass along to him that would be great.

[ChromatographyDude]

Have you tried the 2nd filament?

[LWC]

Yes when I did the PM on the MS I replaced the filaments.

[Ron]

I noted in the posting that when you changed the column the problem went away, but returned in a few days. That leads me to believe the problem is not in the mass spec, but in the column or sample introduction system. What samples are you running, and what sample introduction method are you using?

[LWC]

I have tried 2 different ZB 624 colmns and the problem is in both of them. When the tech was here he put in a DB 5 and "shot" a test solution which passed specs.
I am running EPA method 524.2.
We have tried both manual syringe injections and injections from our purge and trap, it doesn't make a difference the results turn out the same.

[AICMM]

LWC,

I am not terribly useful because I have not played with Shimadzu. However, a couple of comments. First, only work with direct injections. Eliminates the purge and trap for now. If need be, evaporate IS standard in an empty VOA vial and shoot the headspace. Second, is your tune basically stable? That is, if you go in today, will it look basically like yesterday. If so, focus should probably be on the column or inlet. Third, please post your flow conditions, including split ratios and column dimensions. As part of this, I would manually check the flow out the split vent (and septum purge if so equipped.) Finally, you might think about going to a different column, not a ZB 624, as a troubleshooting tool. Easier said than done typically, but considering what you have posted, it may be worth a try.

It is interesting to note that the lens stack was backwards and you could still tune the thing.

Best regards.

[LWC]

Yes the tunes are reproducable, even when the stack was in backwards unbeliveably.
Column 624 60 m x .25 x 1.4
Split 10:1, to see anything. If we increase it to our normal 90:1 split you see nothing.
Flow 1mL/min
The tech that was just here said the flows seemed okay, but I didn't personally watch him do it.

[i.horsting]

We have a shimadzu GC/MS QP2010.

How are your vacuumlevels?
What kind of gas do you use?
What kind of inlet?
What kind of injections are you doing? 1 ul or something else.
Can you post your method?
What is your sample?
What kind of liner are you using?
How did you leak check?
What kind of ferrules are you using?

[LWC]

Our lower vacuum cord is broken so all I have is the high vacuum at 6.9 -003
We have "5 9's" Helium
Not sure what you mean by what kind of inlet?
Injection of 1-5uL or purge trap injections
Method:
inj 230
int 235
rate temp hold
45 10
30 135 1
10 185 3
10 214 1.67
40 245 2.32

MS 235 scan 35-200mZ
sample 524.2 analytes
liner from Shimadzu SPL-17 split
leak detective and check pressurse while dropping flows
ferrules from shimadzu according to manual

[Riikka]

Hi

I don't know if its appropriate to mess with your topic, but I'm also having problems with our GC17A MS QP5000 and the sensitivity, and was wondering if someone could help me out.

Some time ago it was operating like a dream, but then the second filament was broken. We changed the filaments and cleaned the ion source just to be sure (which I guess was the wrong idea...) and because the instrument had been run for long time.

After changing the filaments we had problems with several leaks, but I'm pretty sure those are now under control. BUT the problem is that the sensitivity was totally lost (if I see about 80 peaks with GCFID -> 14 peaks with GCMS, before it was the 1:1 or even more with MS) + background is very high (~700.000). All the values&conditions are the same than before changing the filaments except I've replaced the column with a new DB1-MS!

The autotune report looks ok, but in the mass pattern adjustment the 614 peak is way over 10. H2O peak is higher than N2 when monitoring for leaks. What is a normal intensity for water if the instrument has been on for just a couple of days after ion source was cleaned? The 28 peak is about 1 million, when det. gain is 1,35 kV.

I've never done tuning manually, just used the autotune. Does it help to change the RF gain or other parameters manually or is the problem somewhere else than in the tuning? What about the electron multiplier, how do I know if it's ok or not?

Shimadzu's instruction books are from should I say where...

Thanks already in advance for your kind help!

[LWC]

You said you replaced the column, were the old and new columns exactly the same deminsions?

Our problem ended up stemming from a bad pirani gaguge and split valve in short our flow control was inaddequte with our 60m column, but when using a shorter one everything worked.

Our tunes Intensity calibration coeffiecnt 614-23.4
If your water peak is high, I would say you still have a leak in the system, is it possibly your MS door is not sealed properly. We had an issue a while back also getting the o-ring to sit in the groove.
When I opened up our system and pumped back down it was usually the next morning before the base line looked decent. It would also take several blank runs to get it completly blank.

[Riikka]

Yes, the columns are exactly the same. I was thinking if the old one was somehow ruined and wanted to search for the sensitivity and lower background with a new column.

Water peak is quite high and we did checking for the leaks with IPA monitoring IPA peaks at the peak monitor. -> no leak according to that in the door or the MS nut.

I've run empty runs + standard compounds but the background is not getting lower.

I've tried tuning manually but with my experience it did not show me any improvements... I called the Shimadzu service guy and he suggested that I should carry another GC next to MS and see if it is the GC that is causing the background. I find that quite laborous... Maybe I'll try tuning when column is removed and nut whole plugged...

What do you usually do if the background is high and sensitivity bad?

[AICMM]

LWC,

If I read this right, a high vacuum reading at -003 is not nearly low enough for a quad. You need to be down in the -005 range (below 5 E-5 at a minimum.) Otherwise you will have ion neutral collisions and the whole thing goes to naught. This would certainly affect your tune.

If you don't see a huge air peak, then my next suspicion would be way, way too much flow getting to the MS somehow. If you do not see an air peak, I would shut down the MS and actually measure the column flow at the end of the column.

Best regards.

[Ron]

The vacuum is measured in pascals, not it torr. The high vacuum is fine.

[LWC]

Another thing I kept having issues with was the injection port, it took several attemps to get all the "right parts" before I would go leak free. I guess Shimadzu's QC for parts is not that good, b/c I don't have that kind of problem with my other instrument. So maybe disassemble the injection port and put it back to gether with new parts, it can't hurt to try.
I also saw no leak indicated around the MS door, but as soon as I would put down, and turned the screw the door would pop open, again a new thing even for that instrument.