The Helium Crisis

November 7, 2012

John Hinshaw answers more LCGC readers’ concerns sent to the CHROMacademy forum following the recent CHROMacademy webinar on the helium shortage.

John Hinshaw answers more LCGC readers’ concerns sent to the CHROMacademy forum following the recent CHROMacademy webinar on the helium shortage.

I am using only hydrogen and air for my TCD (thermal conductivity detector) and FID (flame ionization detector) GC (gas chromatography). For the FID the makeup gas choices are helium and nitrogen only. What should be my makeup gas hydrogen flow rate if I enter helium?

The hydrogen makeup gas flow that should go through the FID or TCD for is the same as for helium or nitrogen. It would appear that you are using an EPC (electropolarization chromatography)-type of pneumatic controller and can only choose helium or nitrogen as the makeup gas type. You have two choices, to either (1) use nitrogen or helium makeup gas or (2) enter a lower makeup flow rate, measure the actual flow rate and then adjust the entered flow so that the desired flow is achieved. Pre-calculation of the entered flow is not advisable because the type of flow controller in use will dictate what the effect is of entering a makeup gas type different from the gas in use. Also bear in mind that for a FID the sum of Carrier + Makeup + Fuel hydrogen must equal the required total hydrogen flow through the detector at all times.

What tubing do you recommend for piping your hydrogen to the GC?

Stainless steel is preferred, especially for long-term installations; however, new copper tubing is also acceptable. Any tubing used for GC must be properly cleaned to chromatographic standards.

Is there a way to incorporate hydrogen safely into a mobile laboratory situation without the use of a hydrogen generator? Safety shut-offs or other options for the carrier gas line?

This same situation would arise as if you were using a FID in a mobile laboratory with a hydrogen pressurized tank supply. Ventilate any hydrogen exit flows to the outside, restrain tanks firmly to prevent shifting during transport; use a loop in each gas line to relieve stress; install a hydrogen detector in the lab space and install another leak detector in the GC oven. If possible, connect the alarm relays to a cut-off valve at the low-pressure output side of the gas tank regulator.

How do you get rid of the hydrogen gas coming from the split vent, purge vent and TCD detectors?

Optionally vent these exit flows to the outside with a vent snorkel or route them into an exhaust hood.

So there really is no safety need to route the split and purge flows to an exhaust hood? My company is likely insist on that.

Does anybody just remember the Hindenburg fire in 1937? A layered safety scheme is best as even if the hydrogen flow is routed to an exhaust hood, what happens if the hood fan fails? The same question applies dependent on room ventilation. For better security install a hydrogen detector set to 1% (less than half the explosive limit) that will turn off a valve at the hydrogen tank regulator output and keep it off until it is manually reset. You are then protected if the power, HVAC or exhaust fan fails or if a potentially dangerous level of hydrogen accumulates.

Can hydrogen be used in the analysis of VOCs by purge and trap methods?

Hydrogen is not recommended for purge and trap methods because of hydrogen accumulation in the enclosed purge vessel gas space.

Are there recommendations for using hydrogen as a carrier gas for thermal desorption techniques?

It is not recommended for thermal desorption because of the possible hydrogenation of samples at normal desorption temperatures, especially in the presence of metal desorption tubes.

What about the generation of hydrochloric acid from dichloromethane solutions?

Dichloromethane solutions are not recommended for use with hydrogen carrier gas. This is because hydrochloric acid can be generated in certain situations where liquid solvent is present at elevated temperatures with the hydrogen carrier gas.

According to one vendor, use of hydrogen as a carrier can be a problem when using methylene chloride as a solvent and injector temperatures greater than 275 Centigrade because of the potential for forming hydrochloric acid. Any thoughts on this?

Please see previous question.

We're using our GC–MS with a head space sampler. What do you think about translating to hydrogen for vials pressurization in the head space?

Pressurization of headspace vials with hydrogen is not recommended, as it will create an enclosed pressurized volume of flammable gas. Hydrogen carrier gas can be used with headspace samplers if a separate helium or nitrogen pressurization gas can be used.

What issues would you expect when using an static headspace sampler?

No issues are expected if the hydrogen carrier gas is used with a non-flammable pressurization gas.

Tony Taylor from CHROMacdemy answers some of the other questions sent to the CHROMacademy Forum on this topic:

In our lab we use helium, which is expensive, and so I want to look at changing to hydrogen, however this would render the EI libraries useless - wouldn't it?

There are three points to note here. Firstly, hydrogen is approximately half as viscous as helium, making it slightly more difficult to pump away for high vacuum equipment. Secondly, hydrogen shows minimum plate heights at higher linear velocities than helium. Thirdly, when the outlet pressure drops to almost zero (i.e. a vacuum) the flow of hydrogen into the instrument may be larger.

A larger volumetric gas flow into the instrument when using a direct interface will mean a lower vacuum level is attainable. This may increase the number of background molecules which can collide with the ions formed, leading to a potential reduction in sensitivity and a change in the relative abundance of ions within the mass spectrum.

This all being said, if we use smaller internal diameter columns (0.15 mm, 0.18 mm or 0.20 mm), higher linear velocities are achievable at lower volumetric flow rates. We can maintain the vacuum levels within the system and there should be no wholescale changes in the appearance of the spectra.

As John Hinshaw remarked during the webcast, hydrogen is a reactive gas whereas helium is not and there is always an exception to the rule that states that we do not typically see ion/carrier gas reactions in GC–MS. One that regularly appears under discussion in the forums is the dehalogenation of chlorinated compounds.

If you enable ‘gas saver’ mode on the GC you can drop the split demand significantly, right?

Using a ‘gas saver’ GC mode reduces the split flow after the injector is cleared of sample components and does indeed allow one to save gas by reducing the flow through the split line from, say 200 mL/min to a much lower value, for example 10 mL/min. While maintaining some split flow is desirable to flush the inlet and reduce baseline noise, a considerably lower flow than that for use during the injection phase of the analysis is usually acceptable. This is a great way to conserve helium.

The hydrogen will collect in pockets near the ceiling, won't it?

Hydrogen is less dense than air and will collect at the ceiling, and this is where most folks fit their hydrogen detectors so this is not an issue.

With flow control possible, doesn't it make more sense to optimize flow instead of pressure?

Yes. Most instruments allow the user to select a desired flow, linear velocity or pressure. In temperature programmed analysis constant flow, and in some cases, constant linear velocity can be selected. When changing from helium to hydrogen, the calculator mentioned during the webcast will allow us to translate the initial pressure, linear velocity or flow and we can program our EPC units to adopt these initial settings from where they will calculate the required pressure to maintain constant flow or linear velocity during the temperature program. You can find the method translator at:

Is there an optimum column bore for high split flow hydrogen methods?

This really depends upon your objective. If you want fast analysis, the highest linear velocities are attainable on the smallest internal diameter columns. These columns also give high plate numbers even when relatively short column lengths (10–20 m) are used. Smaller internal diameter columns also require relatively high head pressures to attain the required linear velocity, which makes for more reproducible methods where the EPC units are not trying to reproduce flows at very low applied head pressures.

By keeping the phase ratio constant, the resolution of methods should be preserved. We use 0.18 mm, 0.20 mm and 0.25 mm i.d. columns pretty successfully with hydrogen carrier gas.

Can you please explain again what the concerns were converting to hydrogen as carrier gas with a megabore column?

With megabore columns, the head pressure required to achieve a given carrier linear velocity may be relatively low . A 20 m x 0.53 mm column running at 40 cm/s may require a head pressure of as little as 2 psig. EPC units don’t run so reproducibly at these very low pressures and this can cause reproducibility issues.

Any suggestions on using hydrogen with a packed column? I've been dropping the makeup to keep column plus makeup to be about 45 mL/min.

Yes, use nitrogen! Nitrogen tends to give the best efficiency at the linear velocities typically required for packed column work.

If you’re running micro-packed columns or are stuck with hydrogen, you’re doing the right thing in managing the makeup flow to achieve a total hydrogen flow of around 30–50 mL/min, which is optimal for most manufacturers instruments. Some instruments even have a facility to automatically adjust the makeup flow to achieve a constant column plus make up flow during temperature-programmed operation in constant pressure mode.

How about just not using a makeup gas?

It’s certainly an option .You may not get an optimal response and you should check carefully for deviation from linearity at higher analyte concentrations; however, if the method is fit for purpose you won’t be harming the instrument at all.

We have an FID detector and we use helium as carrier gas. Are you saying we would still need helium or nitrogen as a makeup gas?

Yes. Using hydrogen as a makeup gas usually swamps the detector and alters the stoichiometric ratio of fuel (hydrogen) to oxidizer (air) gases causing a reduction in detector sensitivity. Consider helium, nitrogen or no makeup gas.

Does the ‘hydrogen wear’ effect (metals become brittle when saturated with hydrogen) endanger turbo pumps? Are oil diffusion pumps preferable with hydrogen?

You know I’ve heard this suggested several times and yet I don’t really have a definitive answer. Perhaps a manufacturer representative reading this post will help us to understand a little more (if they have studied this phenomenon). I do know that the use of copper tubing is not recommended when using hydrogen carrier gas because of this phenomenon.

Please comment on the use of a hot cathode ionization gauge to measure MSD pressure when using hydrogen as carrier gas.

My understanding is that most Pirani gauges are in sealed bulbs these days with a permeable membrane used to measure the background current (and hence vacuum). I’m presuming therefore that they will be safe. However, as always when using hydrogen and GC–MS I would very strongly recommend that you contact your supplier to get their instructions before making the switch.

Can you use wider bore capillary columns with nitrogen with GC–MS?

Yes. Nitrogen’s viscosity makes it relatively easy to pump away and achieve a good vacuum. However, I’m not sure that I’d want a whole bunch of nitrogen around. Remember there will be many nitrogen molecules per cm3 even at the best vacuum levels achievable by bench-top instruments. It’s a reasonably reactive gas and one might not easily predict the background/analyte ion reactions that occur in the MS ion source and analyser, which could produce possible anomalies in the mass spectra obtained compared with the less reactive helium.

We normally swap septa without cooling down the inlet. Can we still do that when we switch to hydrogen?

I would not see this as an issue. If you turn off the split flow or reduce your instrument total flow to a low value while carrying out the operation, the amount of hydrogen liberated will be minimal and will not reach the explosive limit of 4% in air. Please ensure that you are in compliance with any local safety rules however and you should perhaps consider fitting a snorkel above the instrument to vent any hydrogen released during this operation or which comes from the split vent or any FID detectors that you might be using.

You mentioned that carrier gas should show a high efficiency over a wide range of linear velocity for temperature programmed GC. But is this really required if you work in constant flow?

Yes. If you want to perform fast separations, high initial flow/linear velocity can push the analytes through the column at the required rate. It’s not so much a case of high linear velocity being developed at the top of the GC temperature programme; it is more that high flow/linear velocity can be used from the beginning when running ‘fast’ or ‘high throughput’ methods and still get very highly efficient peaks.

Please post any other queries you may have regarding the helium shortage on the Chromacademy Forum: ref="">