High water % on tune with Hydrogen

[markf]

Hey guys,

I'm a gc/ms operator for an environmental waste treatment company.
We have a gc/ms, Agilent 6890N/5975B used mainly for EPA methods 8260 and 8270.

About a year ago we switched to hydrogen from helium.
Before that, tune water % was negligible.

When we switched to using a hydrogen generator, we were warned that water % would increase. We set ourselves a "maximum 15% water" to work with.

With time, the water % rose steadily, we investigated and thought that it was due to maintainance needed.

We changed the dissicants on the gas line, changed the Agilent Big Hydrogen Trap...nothing seems to help, the % stays around 20, sometimes even rising to 25%, then decreasing to around 20% again after thermal conditioning.

Furthermore, the number of peaks on the tune printout has risen as well, before hydrogen it almost never rose beyond 150, now it's ~300.
We clean the source routinely.

please advise.

[mckrause]

A quick question. Is this a turbo system?

One of the drawbacks of running hydrogen is that you literally eat your source, thus creating WAY more surface area for your water and other polars to hide out. I'd start with a really good, deep cleaning (polish, not just alumnia paper) and a long bake in an oven. If you have a vacuum oven even better.

If you have a turbo system you may have sputtered away your blades and are in need of a rebuild. Again, a drawback of H2, not to mention the decreased pumping efficiency for turbos of H2. If you're going to run H2 you're way better off with a dif stack.

[mckrause]

One additional thing. If you're running 8260 then you're running purge and trap. There is NO WAY I'm running purge and trap with hydrogen; it's just a really good explosion waiting to happen.

[markf]

1) It's a system with a turbopump, yes. I don't think it messed up the pump though. What other symptoms would I need to look for if that's the case?

2) I'm not running purge&trap, I use headspace for VOCs.

[duck duck goose]

You could try removing your column from the MSD and blocking the transfer line off with a blank nut. Pump the system down and run a tune to see if your water comes down. If it does not, you may have a leak in the MSD, if it does, you know your supply gas isn't good. I guess your O2 would be high too if it was a leak, but that is one way to put another nail in the coffin that you have a bad supply gas.

[James_Ball]

Another thing to consider is that any oxygen in the system could be reacting with the hydrogen to form water, which would increase the water % when looking at a tune.

Make sure you have the proper magnets on the source. 5975s have two different magnets used on the source, one is hydrogen compatible the other is not. 5973s were all compatible but there was a change at some point with the 5975.

I have yet to be able to get BFB to pass criteria using hydrogen carrier for 8260, but I have been able to get DFTPP to pass for 8270.

As for the larger number of peaks in the tune, you will definitely see that happen as hydrogen tends to clean out a system and for months it will bleed out everything that accumulated in the flow path while using helium. Once it is clean though, if you switch back to helium, you will see the cleanest spectra you ever saw

One thing I have seen posted before when switching to hydrogen is to set the instrument to scan overnight with the source temperature elevated and let the hydrogen aggressively clean the source and analyzer. I have not tried it yet myself but it could possibly work. If you search you will find many of us have been trying to make the switch, but have also encountered a ton of problems getting it to work properly.

[markf]

*sigh*

Just wanted to update, a month later...not much success...

Getting 25-30% water on my tunes, numerous solutions have been thought of and tried, such as calling for maintainance on our Hydrogen generator (Leman Instruments), didn't help at all, but maintainance was due anyway.

As a last resort we're trying to order some hydrogen tanks to see if maybe the generator just isn't "good enough" for our uses.

[James_Ball]

Strange that you have high O2 and H2O but low to not existent amounts of N2 and CO2.

What is the other large peak there in the 30 mass range?

If you have oxygen present I guess you could be forming water with the hydrogen in the presence of the hot source.

Have you tried an indicating oxygen trap in the carrier line just before the instrument to be sure you are not getting oxygen impurities from the generator?

[markf]

There's no indicating oxygen trap, there's an indicating silica dryer trap and there's an Agilent Big Universal Hydrogen purifier trap (RMSHY-2) whose main job is to filter water and oxygen, from what I understand.

Also, how can I be sure of the identity of peak in the 30s range?

[serga01net]

As a last resort we're trying to order some hydrogen tanks to see if maybe the generator just isn't "good enough" for our uses.

It is a good idea. High oxygen and water are coming from the H2 generator, I suppose. Part of oxygen reacts with H2 on glowing filament producing high water signal. Maybe, the separating membrane in H2 gen is in poor condition.

[James_Ball]

There's no indicating oxygen trap, there's an indicating silica dryer trap and there's an Agilent Big Universal Hydrogen purifier trap (RMSHY-2) whose main job is to filter water and oxygen, from what I understand.

Also, how can I be sure of the identity of peak in the 30s range?

If the mass is 36 then it might be water (H2O)2 dimer.

If you replace the column with a blank ferrule and scan with no carrier flow it will definitely tell you if you have a leak at the MS or if it is contaminated carrier gas.

If the hydrogen generator is passing oxygen into the carrier stream, then you may have exhausted the oxygen trapping component of the universal trap and it is trapping everything but oxygen now. If so you will probably see a lot of column bleed if you heat the column.

[markf]

There's no indicating oxygen trap, there's an indicating silica dryer trap and there's an Agilent Big Universal Hydrogen purifier trap (RMSHY-2) whose main job is to filter water and oxygen, from what I understand.

Also, how can I be sure of the identity of peak in the 30s range?

If the mass is 36 then it might be water (H2O)2 dimer.

If you replace the column with a blank ferrule and scan with no carrier flow it will definitely tell you if you have a leak at the MS or if it is contaminated carrier gas.

If the hydrogen generator is passing oxygen into the carrier stream, then you may have exhausted the oxygen trapping component of the universal trap and it is trapping everything but oxygen now. If so you will probably see a lot of column bleed if you heat the column.

The universal trap was changed about a month ago, I'd like hope that the situation is not that bad as to fill up a new trap in such a short time...

A stupid question here, but, when talking about 20% water in tune, it's water abundance in relation to peak 69, right?
I have a gc/fid on the same H2 gas line and the FID has no trouble lighting up, I thought that constant water in the carrier gas (And the fid combustion gas) would interfere with the detector.

Anyway, when working with EPA8270 we're heating the gc to 300 and have not seen any more bleeding than the ordinary (Baseline rising and few siloxane peaks).

[James_Ball]

It is 20% of peak 18 versus 69 from the PFTBA, but would only amount to low ppm concentrations of water in the system overall. You probably wouldn't see any effects on the FID since an FID can take a microliter injection of water and remain lit, where that would be high percentage levels of water in the carrier stream at the time it passes through the FID.

If it were an air leak then nitrogen would be several times more than oxygen (about 5x) in that last tune report, but since nitrogen is 5x less than oxygen it would be safe to say it isn't and air leak. CO2(m/z44) is very low along with Argon(m/z40) which also suggests it isn't and air leak. About the only way to get oxygen in the system without an air leak would be in the carrier gas. The mixing of the oxygen and hydrogen is most likely the source of the water. Definitely need to look into the hydrogen generator I would think at this point.

[markf]

Little update:

Water % is now 50-60, and we changed to hydrogen tanks 2 days ago, no change observed...I'm puzzled - if it's not the carrier gas - what coud it be?

The number of peaks on tune has risen to about ~500, which is also worrying me.

I've called a serviceman, he said that he will deep clean the ion source and he thinks that there will be an improvement after that. To be honest, I'm skeptic.

[James_Ball]

The increase in peaks could be the hydrogen cleaning all the junk out of the gas lines and inlet if the instrument was run for a long time on helium, I know it happened to mine when I tried hydrogen carrier.

One site I read once suggested when switching to hydrogen, elevate the source temperature to the highest setting and turn on the filament and let it scan without tune gas all night to help burn out any contamination. I never tried it because I didn't want to run a filament that long, but who knows.

Does the water amount change if you run the inlet split ratio up about 100:1?