H2 analysis byTCD/MolSieve/Argon

[cmoral]

Hi,

I also have a question about analyzing for hydrogen using a TCD and argon. I have set up my ancient HP5890 Series II GC (works great by the way) with a TCD and have optimized it for O2, N2 using helium and a mol sieve column. I presume I can use it for other permanent gases. I can only detect pure hydrogen with a very poor response and as a negative peak using helium which I expected so decided to switch to argon. I don't have a problem with negative peaks. My software can integrate them so that's not the problem.

The problem is that I found as soon as I switched the carrier gas to argon, I could not increase my column head pressure at all from essentially 1 PSI, which seems to also be a baseline if I were to shut down the helium. It had been 5 PSI for helium. I know argon is a heavier molecule and runs would take longer but this was a total surprise to get nothing registering or be able to adjust.

I also do not register total flow for the split valve or the the septum purge although no problems with helium. I can measure a flow through the column (I think unless it was some residual helium or something) depending on how much I open up the reference flow.I don't have electronic flow sensing on this instrument and not sure if there is something in the software (using Agilent Chemstation ver B.01XX) that I need to adjust for the argon.

Can anybody give me a reason why? I have no experience with argon...plenty with hydrogen and helium.

Thanks.
cmoral

[pwm]

Ive run a similar setup and get very high pressures with Ar, >>1psi. I dont know how the same system would compare on He, but I imagine in the same ballpark. 1 psi, sounds like you popped something loose while switching gases. Though you say replumbing with He all returns to normal?

[cmoral]

Thanks for your reply. Yes, it all returns to normal with He but as soon as I switch to any other gas, everything just completely drops to 0. So don't think I popped anything off. Am thinking of switching to 92% He/8% H2, which was suggested as a good carrier for determining H2 and that might work fine because most is He but it would be nice to know why I have the problem with other gases. Can't specifically see anything to change in the software parameters.

[benhutcherson]

Just to sort things out, are you using a 5890 with the standard manual pneumatic control, or do you have the EPC?(gauge and three knobs on the left panel vs. no gauge and two knobs). With the EPC, be sure you've changed the carrier gas, although I'm 99% sure argon isn't one of the options, and that still shouldn't affect your ability to change pressure.

In general, though, if you can't raise your head pressure it's worth checking all the obvious ones. Make sure the column is installed correctly, and the column nut is apopriately tight. Make sure you have a good O-ring on the liner, and that the weldment nut on top of it is tight. Make sure the septum is good and also tightened correctly.

Also, the plumbing on these is a bit counter-intuitive. IIRC, the knobs on the manual pneumatics are head pressure, split flow, and septum purge, but if any of these are shut off or two low you can have issues raising the head pressure. With the EPC, having the "total flow" knob set too low or two high can also wreak havoc with it.

As a side note, I usually do hydrogen with nitrogen. Nitrogen will still allow hydrogen to give a nice, strong negative signal but is a lot less expensive.

[cmoral]

Hi Ben,

Am using an HP5890 series II with no EPC control just two black knobs for ref and makeup gases....i think...sorry, am shut out of the lab due to covid so cant totally remember if it's two or three knobs. the only thing I changed was the connection from the helium gas line to the argon line. I tried it first with nitrogen as well and same problem. But as soon as I switch back to the helium line (set to 40 psi) no problem so it can't be leaks anywhere in the lines.

It's like as soon as I connect up the argon or n2 the column head pressure just doesnt go up even with valve completely open and that prevents flow to the column. Is it possible the problem is with the gauge not set to allow n2 or argon pressure?. The max n2 line pressure coming from our wall line is 20 psi which might be too low but the argon from a cylinder definitely set at 40 psi...even tried 60 psi and even 80 I think and zippo.

Can't get back into the lab for awhile but if you have any other insights that would be appreciated. I could try when I get back in. Thanks.

[benhutcherson]

This SHOULDN'T make a difference but may...I've never played with it but this may need to be changed with a different gas.

5890s w/o EPC(and/or a mass spec) typically have a regulator panel "hidden" inside the instrument. If you open the door on the left hand side, or remove the left panel completely, you should see it. It's an upright panel with three gauges and a knob below each one. AFAIK(never really looked into it) this is a single stage regulator that serves as a "check" before gases go to the final regulator on either the inlet or regulator. At least on the 5890 FIDs I've used, the regulators are labeled for helium, hydrogen, and air. I've never worked on a 5890 TCD so don't know what you'd find in there, but presumably there's at least one marked helium. You may need to up the pressure on the "helium" gauge when plumbing in a different gas.

(due to COVID, I'm also away from the lab and even though I have two working/in use 5890s can't access them either right now).

[Consumer Products Guy]

Thanks for your reply. Yes, it all returns to normal with He but as soon as I switch to any other gas, everything just completely drops to 0. So don't think I popped anything off. Am thinking of switching to 92% He/8% H2, which was suggested as a good carrier for determining H2 and that might work fine because most is He but it would be nice to know why I have the problem with other gases. Can't specifically see anything to change in the software parameters.

We didn't have need often to assay for hydrogen. But we did get a cylinder of 92.5% helium and 7.5% hydrogen to use as a carrier gas. And used TCD, also on a 5890 or 5840 GC (predecessor to 5890).

[cmoral]

Thanks, Ben. Funny how I have worked inside this instrument numerous times for 25 years and yet cannot remember all the workings on the inside. I am beginning to wonder if it is something with the regulator on the inside but have to wait to get back into the lab in a month (hopefully) or more (boo!) to look. Once I can look inside, I'll let you know what I find. If I can use N2 and not argon that would be great.

[benhutcherson]

With manual pneumatics, I can't really imagine any issue with using argon.

The EPC shouldn't have an issue either aside from the fact that I think N2, H2, and He are the only options it gives you when calculating flow rate.

[cmoral]

I agree with you, which is why this is totally confusing me. But once I switch the inlet line gas line from He too N2, Ar, and even H2 into that gauge, I get zippo and once I change back to He, it immediately goes back up. It's a puzzler.

[benhutcherson]

You honestly have me stumped also.

H2 sometimes shows issues both with leaking where other gases won't and setting a desired pressure because its viscosity is so low, but N2 and He are similar in viscosity(that was a weird one to me when I looked it up, as I would have expected N2 to be a whole lot higher) and without checking I'm guessing Ar is similar also.

When change gases, where in the GC are you swapping the fitting? The "normal" set up I've seen on installed instruments is to run it either to the gas manifold/regulator panel if installed, or otherwise to run it directly to the front panel pneumatic control. With that said, a lot of instruments now in circulation will have a short "pigtail" running out the back of the instrument to allow for connecting and disconnecting without pulling the side panel.

[cmoral]

The line runs directly into the front panel pneumatic control, which is why I am puzzled. I checked the gas flow out of the line to make sure its not blocked anywhere else (its a direct line from the Ar tank or N2 output so didn't expect any block) before putting it into the control fitting. since we use N2 for our other GC and other flow set ups, I wasn't surprised that I wouldn't necessarily get anything as our max pressure out of the line is 20 PSI and dividing it up might lower than enough to be an issue but directly out of a H2 or Ar tank where I can crank up the pressure shouldn't give me a problem.

Wondering, if for some reason, the actual knob that controls the pressure into the gauge is somehow set only for He. I know that if I mess up the control knobs for the FID gases when replacing or cleaning them, I get flows out of whack...i.e. H2 too high, air flow too low (learned that the hard way!).

I really appreciate your help on this and am grateful if you actually find something I haven't figured out but until I can get back into the lab, unfortunately, I can't check anything for you.

[benhutcherson]

As such things go, 5890s are pretty primitive(doesn't change that they're one of the most solid GC designs of all time and even current Agilent instruments are really just refined/updated versions) and the manual pneumatics are "dumb" devices. They don't care what gas they are being fed just as long as there isn't something in the flow path that's incompatible with the gas.

In fact, even in an 8890 the only super relevant thing about specifying the correct gas is so that the flow rate is correct. You should tell it the correct gas, but it will still work(work in the sense of providing flow to the column) if not.

I did pull out my set of 5890A maintenance manuals and unfortunately they weren't a ton of help on this.

I don't want to state the stupidly obvious, but just to make sure-you increase the head pressure by turning the knob CLOCKWISE and not COUNTER-CLOCKWISE. You are doing this, correct? Sorry if this comes off as patronizing, but I know I've made plenty of stupid mistakes in my life by confusing that(my best/worst one was the time I replaced the headgasket on a car thinking it was breached between two cylinders all because the spark plugs were hooked up wrong. I stood there and said repeatedly "1-3-4-2 counterclockwise" while going clockwise around the distributor-not my brightest hour or rather 6 hours).

[cmoral]

Actually, you're not being patronizing. I have done enough of those things myself to realize that I can do these dumb things again. And, yes, I did turn it both ways to make sure and still zippo. Let's just wait until I get back into the lab and then I will repeat everything stepwise and get back to you with the results. There is an explanation for this.....somewhere....somehow.