There are two main types of purifiers used for gases: getters and membranes. Both can work for you in almost any application, and we make both types at REB Research – for hydrogen purification mostly, but sometimes for other applications. The point of this essay is which one makes more sense for which application. I’ll mostly talk about hydrogen purification, but many of the principles apply generally. The way both methods work is by separating the fast gas from the slower gas. With most getters and most membranes, hydrogen is the fast gas. That is to say, hydrogen usually is the component that goes through the membrane preferentially, and hydrogen is the gas that goes through most getters preferentially. It’s not always the case, but generally.
Our getter beds for use with inert gasses have two chambers; one is high temperature to remove water, nitrogen, etc. and one at lower temperature the remove H2. The lower temperature bed can be regenerated.
Consider the problem of removing water and similar impurities from a low-flow stream of helium for a gas chromatograph. You probably want to use a getter because there are not really good membranes that differentiate helium from impurities. And even with hydrogen, at low flow rates the getter system will probably be cheaper. Besides, the purified gas from a getter leaves at the same pressure as it entered. With membranes, the fas gas (hydrogen) leaves at a lower pressure. The pressure difference is what drives membrane extraction. For inert gas drying our getters use vanadium-titanium to absorb most of the impurities, and we offer a second, lower temperature bed to remove hydrogen. For hydrogen purification with a bed, we use vanadium and skip the second bed. Other popular companies use other getters, e.g. drierite or sodium-lead. Whatever the getter, the gas will leave purified until the getter is used up. The advantage of sodium lead is that it gets more of the impurity (Purifies to higher purity). Vanadium-titanium removes not only water, but also oxygen, nitrogen, H2S, chlorine, etc. The problem is that it is more expensive, and it must operate at warm (or hot) temperatures. Also, it does not removed inert gases, like helium or argon from hydrogen; no getter does.
To see why getters can be cheaper than membranes if you don’t purify much gas, or if the gas starts out quite pure, consider a getter bed that contains 50 grams of vanadium-titanium (one mol). This amount of getter will purify 100 mols of fast gas (hydrogen or argon, say) if the fast gas contains 1% water. The same purifier will purify 1000 mols of fast gas with 0.1% impurity. Lets say you plan to use 1 liter per minute of gas at one atmosphere and room temperature, and you start with gas containing 0.1% impurity (3N = 99.9% gas). Since the volume of 100 mols of most gases a these conditions is 2400 liters. Thus, you can expect our purifier to last for 400 hours (two weeks) at this flow rate, or for four years if you start with 99.999% gas (5N). People who use a single gas chromatograph or two, generally find that getter-based purifiers make sense; they typically use only about 0.1 liters/minute, and can thus get 4+ years’ operation even with 4N gas. If you have high flows, e.g. many chromatographs or your gas is less-pure, you’re probably better off with a membrane-based purifier, shown below. That what I’ll discuss next.
Our membrane reactors and most of our hydrogen purifiers operate with pallium-membranes and pressure-outside. Only hydrogen permeates through the palladium membrane.
The majority of membrane-based purifiers produced by our company use metallic membranes, usually palladium alloys, and very often (not always) with pressure on the outside. Only hydrogen passes through the membranes. Even with very impure feed gases, these purifiers will output 99.99999+% pure H2 and since the membrane is not used up, they will typically operate forever so long as there is no other issue — power outages can cause problems (we provide solutions to this). The main customers for our metallic membrane purifiers are small laboratories use and light manufacturers. We also manufacture devices that combine a reformer that makes 50% pure hydrogen from methanol + steam where the membranes are incorporated with the reactor — a membrane reformer, and this has significant advantages. There is no equivalent getter-based device, to my knowledge because it would take too much getter to deal with such impure gas.
Metal membranes are impermeable to inert gases like helium and argon too, and this is an advantage for some customers, those who don’t want anything but hydrogen. For other customers, those who want a cheaper solution, or are trying to purify large amounts of helium, we provide polymeric membranes, a lower cost, lower temperature option. Metal membranes are also used with deuterium or tritium, the higher isotopes of hydrogen. The lighter isotopes of hydrogen permeate these membranes faster than the heavier ones for reasons I discuss here.
Robert Buxbaum, August 26, 2018
