Neon vs Argon

Both gases are inert, colourless and monatomic, and there the similarity stops. Argon is one of the cheapest industrial gases in existence and is used by the hundreds of thousands of tonnes. Neon costs two to three orders of magnitude more per cubic metre and is bought in quantities that would embarrass an argon supplier.

Everything about that gap comes from one number: how much of each is in the air.

Five hundred times the air for the same litre of gas

Argon makes up 0.934% of the atmosphere, which places it third behind nitrogen and oxygen and well ahead of carbon dioxide. Neon is present at about 18 parts per million. To collect the same volume of neon you have to push roughly five hundred times as much air through a plant.

The economics follow directly. Every cryogenic air separation unit built to make oxygen is already fractionating argon whether the operator wants it or not — argon boils between nitrogen and oxygen and is drawn off a side column. It is a co-product with essentially no incremental capital cost.

Neon is not. It concentrates in the light, non-condensable fraction at the top of the column, and recovering it requires additional equipment that only a minority of plants have. Nor is the price a shortage of the raw material: cosmically neon is one of the half-dozen commonest elements there are, and where exactly it places in that half-dozen depends on whether atoms or mass are being counted and on whether nitrogen is put ahead of it. The expense is a purely local accident. Neon is too light for this planet to have held gravitationally over geological time and too inert to be locked into any mineral, so what escaped never came back.

Argon avoided that fate twice over: it is heavier, and it is continuously replenished. Almost all atmospheric argon is argon-40, produced underground by the decay of potassium-40 and outgassed since the planet formed. The argon in the solar wind is a quite different isotopic mixture, dominated by argon-36. Earth's third most abundant atmospheric gas is, in effect, radioactive exhaust.

The gas found by weighing nitrogen twice

Argon's discovery is one of the best arguments in science for taking a small discrepancy seriously. Two batches of nitrogen — one pulled out of the atmosphere, one cooked up from ammonia — would not agree on their weight, and the size of the disagreement was about half a per cent, which is the size most working chemists of the 1890s would have blamed on their own apparatus and moved past. Lord Rayleigh refused to, and with William Ramsay he cleared an air sample of every component then known until what was left over would react with nothing offered to it. They named it for laziness.

Ramsay and Morris Travers found neon four years later by fractionating liquefied air, along with krypton and xenon in the same run.

Neon's supply chain ran through steelworks

Semiconductor-grade neon has an unusual industrial history. Much of the world's crude neon was recovered as a sideline from very large air separation units originally built to supply Soviet-era steel plants, with the final purification concentrated in a handful of Ukrainian companies. By the early 2020s that route supplied roughly half the world's chip-grade neon.

The reason it mattered so much is that neon has become indispensable to lithography. The excimer lasers that expose modern chips work at 248 and 193 nanometres, on krypton-fluorine and argon-fluorine mixtures respectively, and in both cases the gas fill is overwhelmingly neon by volume with the halogen and the heavier rare gas present only in traces. Neon is the buffer that carries the discharge.

When that supply was disrupted in 2022, neon prices rose by an order of magnitude within months and chipmakers went looking for alternatives, with substantial new purification capacity built in China and elsewhere. It remains one of the clearest examples of an obscure industrial gas turning out to be a chokepoint for a trillion-dollar industry.

Only one colour in a neon sign is neon

The industry named itself after the gas it uses least of. Neon's discharge is a strong red-orange dominated by a line at 640 nanometres, and that single colour is the whole of its repertoire. Every blue, green, pink, yellow and white tube in a shop window is running on argon instead, with mercury and a phosphor coating between them deciding what comes out. Put the two purchase orders side by side and a sign shop is overwhelmingly an argon customer, which is the reverse of what the trade name implies.

The two gases behave differently in a discharge for a concrete reason. Only helium holds onto its outermost electron more tightly than neon does, and argon parts with its own several electronvolts more readily. Argon strikes and sustains a plasma more readily, which is why it is also the standard gas for sputtering, plasma cutting and inductively coupled plasma spectrometry.

Two kinds of inert

On the criterion that matters when you are choosing a gas — will anything bond to it — neon is the hardest case in the table, and argon shows just how narrow the margin between them is. (On the different criterion of how rigid the atom is, it is helium and not neon that holds the record; the neon page keeps the two claims apart, because they are routinely conflated.)

Argon fluorohydride, made by a Finnish group in 2000, is a real molecule with a genuine bond from argon to hydrogen. It survives only trapped in a solid matrix below about 27 kelvin, so it is not a chemistry anyone can put to work, but it is a bond, and it proves argon's electrons are reachable in principle. Neon's cold-matrix species are a different animal. Nothing in them is bound by more than a few kJ per mole, which is contact rather than chemistry. The distance between weakly bonded and not bonded is the whole of the difference here, and argon is on the near side of it.

For the practical purpose either gas is bought for, both are absolutely unreactive and the distinction never comes up. It matters only for the argument about which element is the most inert, where the answer changes with the question.

Where each one goes cold

Argon's industrial life is mostly at high temperature: shielding the arc in welding, blanketing molten titanium and specialty alloys, and above all the argon-oxygen decarburisation step that most stainless steel passes through. Its everyday appearances are quieter — the fill gas in double-glazed units, where it conducts heat less readily than air, and in incandescent and fluorescent lamps.

Neon's specialist niche is at the other extreme, and it is the one duty on this page that argon cannot be entered for at all. Argon condenses at 87 kelvin; neon holds on for another sixty degrees, down to about 27, which drops it into the narrow band between hydrogen and helium. A coolant picked for that band is being weighed against those two, never against argon, and neon's case is that a given tankful of it absorbs far more heat than the same tankful of either — enough to justify a price an argon buyer would think was a typing error. Neon also supplies the lasing transition in the helium-neon laser, whose 633-nanometre red beam was the standard laboratory laser for a generation.

Filling a tube, a weld, or a cryostat

• Shielding a weld or a melt — argon, and cost is barely a consideration.
• Sputtering, plasma etching or plasma spectrometry — argon, which ionises more easily.
• Insulating a window cavity — argon.
• Running an excimer laser for photolithography — neon, as the bulk of the mixture.
• Refrigeration around 25 to 30 kelvin — neon, where the volumetric capacity justifies the bill.
• A tube that must glow red without a phosphor — neon, the only one of the two that does.