GC-MS detector voltage increasing after cleaning...

[KevinBe]

Hi All,

I am absolutely not an expert in GC-MS...

But here is what is happening with out Shimadzu GCMS

One student injected over concentred samples in the GC-Ms and contaminated it...
So I decide to do a full clean up.

I change the septum. The liner, I trim the column on both end and I decide to clean the ion source too.

I also found out a leak in a connector on my He cylinder. I did use a tiny bit of soapy water to find that leak. I quickly fixed it.

After pumping down, I see that I have quite a lot of water... (my board for using soaping water I guess) but within two hours the level of water is back to normal.

I run the auto tune for our combined NCi/CI/EI source and bad luck... the voltage of the detector increased compared to before the cleaning! I cleaned the source again, sonicated it in water, acetone and MeOH. Dried it at 300C... and again the same tuning values!

I have no idea what I have done wrong... help!

Thanks a lot!

I am attaching the tune profile after cleaning

https://ibb.co/zGvzByW

And before cleaning

https://ibb.co/DzYd2VB

[dblux_]

... I have no idea what I have done wrong... help!
Thanks a lot!

IMHO nothing wrong (besides repeating cleaning of ion source).
If we both think of electron multiplier than voltage increases over time.

[KevinBe]

Thanks a lot

Well... the voltage was a 0.9kV and jumped to 1.23kV after cleaning.

The base line for a blank injection of MeOH was at 20 000 counts before cleaning and is now at 60 000 counts after. Isn’t the sign of an issue?

I also tuned the second filament (that we never use) and the values where the same - 1.20kV

[James_Ball]

Thanks a lot

Well... the voltage was a 0.9kV and jumped to 1.23kV after cleaning.

The base line for a blank injection of MeOH was at 20 000 counts before cleaning and is now at 60 000 counts after. Isn’t the sign of an issue?

I also tuned the second filament (that we never use) and the values where the same - 1.20kV

Not sure on a Shimadzu but those values don't sound out of the norm for an Agilent multiplier voltage. Did you bake the source under vacuum at 300C? If it was in an open oven you could have oxidized the source, but under vacuum it should be ok.

Give it another day and recheck the baseline and see if it goes down. What ions are you seeing in the baseline that elevate it up to 60k from 20k?

I normally try to wait overnight after cleaning a source to do anything, as even the best ones take a while to fully equilibrate and bake off any residuals from the solvent used for cleaning and the moisture that gets into the vacuum chamber when it is open.

[KevinBe]

Thanks.

Well Shimadzu has a source made of vanadium i think and it might not oxidise that quickly because it says on their maintenance manual to bake it at 300C in a normal oven. So I did that for 1H, then reassembled it and now it’s a 200C overnight I’m under vacuum.

[KevinBe]

Oh yes, I also changed the pump’s oil... in case that matters...

[KevinBe]

The 207 ion has dramatically increased in my base line! Could it be that the column is dead?!

[benhutcherson]

Let's take a step back and make sure everything is right in your source before looking for other problems.

I'll admit zero knowledge of Shimadzus, but at least in other GC-MSs a dramatic increase in multiplier voltage that quickly(as opposed to the gradual increase that happens with use) is usually a source problem.

Are you sure everything is assembled and attached correctly? Further to that, sources are full of various types of insulators, and depending on the exact design they are made of different materials. Ceramic and vespal are the two I'm most familiar with. A cracked or damaged insulator somewhere in the ion optics can cause a dramatic jump in multiplier voltage.

That would be where I'd look.

[James_Ball]

The 207 ion has dramatically increased in my base line! Could it be that the column is dead?!

With the increase in background at 207 which is normally column bleed, it could be the column but have you injected a known standard that you have historic data on that you can compare peak height or area counts for before and after?

On the tune reports from autotune, do the mass 69 counts agree between one before and one after the cleaning, or is the one after now higher? If it is it could just be that the autotune increased the multiplier voltage a little and it is making everything look different. I know with Agilent when it autotunes, if there is a little water left in the analyzer it can affect the electron multiplier setting and if you tune again the next day it will be slightly different.

[benhutcherson]

I know with Agilent when it autotunes, if there is a little water left in the analyzer it can affect the electron multiplier setting and if you tune again the next day it will be slightly different.

I've been talking a lot recently with an independent/former HP FSE out of Michigan, and have worked a deal(with the boss's approval) to buy a full set of HP 5972 electronics(PSU, sideboard, smartcard) plus he's giving me a 5972 vac manifold and some other odds and ends(including all the parts needed to add CI to a 5972). Once done, I'll in essence have a field-upgrade 72 that says 5971 on the front. There's actually a lot of interesting things that that allows, and I'm hoping to write up the whole process when it's done.

In any case, I was talking to him about how annoyingly long a 5971 takes to "settle" after being vented, even if the vent is only for a few minutes. I told him I normally did maintenance on Friday and didn't even try to tune it until Monday. His response was "Yeah, with a 71 the best thing to do is put it back together, start a pump down, then give it a hug and say 'see you in a few days." Apparently the polished 72 vac manifold cuts the settle-down time by a fair bit.

My experience is almost entirely with HP/Agilent and now Varian. After maintenance, I never clear one for use until at least an overnight pump-down. It can take a while to get everything out of the source. The Varian triple is a bit better since in routine maintenance you only touch the ion volume, rarely mess with any of the optics, and this can be done without venting. Even with that, it still seems to like an hour or two after a clean source is installed to give a good tune.

[KevinBe]

Thanks everyone!

So the voltage seems to become a bit better with the time (it’s getting down but not dramatically...)

I checked the source again, sonicated it in water, acetone and methanol. I checked also the insulators (ceramic) and they are alright.

I now put a new column in (brand new!) and i have a new issue... the baseline just once the column heats up at 100C... the signals are the one of an alkane 57, 71 peaks.

I did forget to wipe the outside of the column that gets into the transfer line with acetone but if this was the issue I should get these ions all the time and not only at >100C of the oven since the transfer line is at 250... maybe I need to change the liners and septum...

I did change the pump oil while I was doing a full cleaning of the GC... and I might have done something wrong!

[James_Ball]

Thanks everyone!

So the voltage seems to become a bit better with the time (it’s getting down but not dramatically...)

I checked the source again, sonicated it in water, acetone and methanol. I checked also the insulators (ceramic) and they are alright.

I now put a new column in (brand new!) and i have a new issue... the baseline just once the column heats up at 100C... the signals are the one of an alkane 57, 71 peaks.

I did forget to wipe the outside of the column that gets into the transfer line with acetone but if this was the issue I should get these ions all the time and not only at >100C of the oven since the transfer line is at 250... maybe I need to change the liners and septum...

I did change the pump oil while I was doing a full cleaning of the GC... and I might have done something wrong!

Those masses can also come from the mobile phase of a brand new column. I like to ramp mine up to the max isothermal temp for a while and let them bake out. You can also get a lot of water and nitrogen coming out from the phase when it first bakes.

When you are sonicating, are you also sonicating the ceramic insulators? If so, those will take a while to loose all the water they absorb even under vacuum and high temp, which will also shift your multiplier voltage. I would say let it pump over the weekend and see how it compares to an older tune. With even a few months use, a multiplier can be expected to have a small increase in required voltage, they are considered a consumable part and have a finite lifetime. Read up in the manual and see what the maximum operating voltage is for the multiplier you are using, once you get to between 50% - 75% of that, then begin to worry about replacing it. Otherwise, let it self adjust and go with it. It is nearly impossible to have the exact same response for a standard over a lifetime of the detector, it will always change over time.

[KevinBe]

Thanks a lot,

So here is what I have done:

I have removed the column, reattached it and this time I wiped the bits that get into the MS and the injector with dichloromethane and acetone.

I also changed the ferrules, the liner and the septum,

I rebaked the column at 280 for a couple of hours and the source at 250C.

And tada.... with a source at 200, a transfer like at 250:

an acceptable vacuum leak (69/28 =23), a bit of water and a decent spectrum! The base line is acceptable even though I have 207 as the major ion. Flat baseline until 250C, after that it increases between 250 and 300. So the new column is still bleeding... (it’s supposed to have a higher limit of 340!)

Still the detector voltage is high... but we went from 1.33kV in the morning to 1.31kV in the night so I hope to see it getting down over the weekend.

Still I have two questions:

1. Is water bad for the source? I mean can I turn on the filament to monitor the decrease of water or is that a really bad idea?

2. When you bake a column, how long do you bake it for and what should be the total flow and the detector temp?

Thanks!

[benhutcherson]

I generally don't go nuts on new columns.

First of all, I buy ultra-low-bleed MS columns. In my experience, they are worth the extra money, and in recent years I see next to no bleed on a new column even with minimal conditioning. With that said, I routinely install carbowax packed columns in an FID with no conditioning and don't see any appreciable bleed, although of course it's difficult to pick out what in the baseline is bleed on an FID.

In any case, I do generally give a new column a quick condition, but that's about it. My usual procedure is to fit it to the inlet, set the flow rate to ~2mL/min, set the oven to either 20º higher than the max for any method that will be run, or in general purpose systems to the maximum isothermal temperature indicated on the tag(note that the isothermal temperature is usually somewhat lower than the program temperature-on a typical J&W HP-5MS it's 325/350º). I will give it a half hour of flowing into the column oven, or in other words not connected to the MS, although if running hydrogen I'd probably want to at least have it running outside the oven. When I need to do it in a hurry, I've been known to start this before the MS has fully finished venting, and just shove the inlet end of the old column into a septum to minimize the amount of air entering it.

[James_Ball]

Thanks a lot,

So here is what I have done:

I have removed the column, reattached it and this time I wiped the bits that get into the MS and the injector with dichloromethane and acetone.

I also changed the ferrules, the liner and the septum,

I rebaked the column at 280 for a couple of hours and the source at 250C.

And tada.... with a source at 200, a transfer like at 250:

an acceptable vacuum leak (69/28 =23), a bit of water and a decent spectrum! The base line is acceptable even though I have 207 as the major ion. Flat baseline until 250C, after that it increases between 250 and 300. So the new column is still bleeding... (it’s supposed to have a higher limit of 340!)

Still the detector voltage is high... but we went from 1.33kV in the morning to 1.31kV in the night so I hope to see it getting down over the weekend.

Still I have two questions:

1. Is water bad for the source? I mean can I turn on the filament to monitor the decrease of water or is that a really bad idea?

2. When you bake a column, how long do you bake it for and what should be the total flow and the detector temp?

Thanks!

If water is over ranging the detector I stop scanning, but once I can see it is just high but not maxing out I have left it on to watch it and see if it falls off. If it does, I turn it off then come back an hour later and check again. When it is 2x higher than 69 I just leave it on and monitor until I see it go away.

If you have the threshold set to zero, then you will see some 207 at just about any oven temperature, there will always be a tiny amount of bleed, simply because the very end of the column in heated at the MS interface. By the time you get to 150c you will begin to see it increase slightly, above 250c and you will definitely see some even with low bleed columns, and over 325c it can increase quite a bit.

What were typical counts for 69 when the multiplier was just below 1kv and what are they now at 1.31kv? A 20v drop is really negligible, I can run two auto tunes back to back and see that much difference or more.

How old is the detector?