Hacker News new | past | comments | ask | show | jobs | submit login
Understanding the science behind the neon shortage (advancedsciencenews.com)
63 points by sillystuff 10 days ago | hide | past | favorite | 17 comments

There were a couple of articles in February that stated 90 percent of the neon used in semiconductors came from Russia and Ukraine, among many other critical supplies for western technology:

"Some 90pc of the world supply of neon, used as laser gas for chip lithography, comes from Russia and Ukraine. Two-thirds of this is purified for the global market by one company in Odessa. There are other long-term sources of neon in Africa but that is irrelevant in the short run."



Also an HN thread back then:

https://news.ycombinator.com/item?id=30457490 (422 comments)

> There were a couple of articles in February

The article that is the subject of this thread was published in March.

So pretty much part of the same series, so to speak.

This is not all bad news - manufacturers are realizing that just-in-time global supply chains are highly fragile and this is spurring quite a bit of domestic investment in production facilities. Helium shortages are as much a problem as neon shortages, but that's driving specialty gas production domestically

> "For example, helium, a critical material for semiconductor manufacturing, has been under considerable strain recently. The inert gas has a thermal conductivity second only to hydrogen, so it is perfect for safely cooling down chips and tools during the patterning process. Helium is also used in some lasers, and ionized helium is a potent etchant. Many analysts describe helium as irreplaceable for semiconductor manufacturing, and the new nodes use more of it than ever... Russia’s invasion of Ukraine further disrupted helium supplies by scrambling the natural gas market; most helium is extracted as a side product from natural gas wells. The current helium shortage is the fourth in recent history, according to helium consultant Phil Kornbluth, and it may be the worst yet."


This article gives me more questions than answers.

- (1) If neon is produced by fractional distillation of air, doesn't that mean it can be produced anywhere? What's stopping someone from building a plant in the US or Europe or China or whatever?

- (2) If neon's a byproduct of nitrogen extraction for fertilizer manufacturing, surely Ukraine's not the only place that produces fertilizer? Are there other existing plants outside the conflict zone that liquify air to get nitrogen for fertilizer, or oxygen for medical or rocket launch purposes, that could be easily modded to extract neon as well?

- (3) Demand is 500k tons per year according to the article and neon is 1 part per 79000 by mass of Earth's atmosphere according to Wikipedia. This means we must be processing 79k x 500k = 39.5M tons of air per year. I don't know enough about industrial gas processing to estimate how big / expensive / many facilities you'd need to process air at that rate?

- (4) It seems like when separating air cryogenically, you'd get not just oxygen, nitrogen and neon, but also CO2 and coldness for free. Once you distill the CO2, couldn't you capture it for environmental reasons or sell it as dry ice or a fire extinguisher?

- (5) You really only care about cooling the air temporarily to make it separate into its component gases. Once it's separated, you let it warm back up again -- but doesn't that waste all the energy you used to cool it down? Since it's not geographically restricted by resource deposits and can be sited anywhere with reasonable infrastructure (adequate electricity and accessibility by pipes, trucks or trains to haul away the gases), can't you put it next to someplace with a constant demand for cooling, like a datacenter, and then pipe the cold gases past the hot servers a few times before compressing them for storage? That way you don't waste all the BTU's you used to cool down the gas.

So my understanding of the situation, is that during the Soviet years the USSR prioritized having a neon stockpile, and the end result was that after the collapse of the USSR, there was so much of a supply glut of neon coming from the former states that it was not possible to compete on (1) or (2), since ex-Soviet factories had the equipment installed and paid for already. And nobody was really betting that there would be an actual full-on land war in Ukraine that would make dependence on this supply glut a problem.

Here's another HN comment with sources: https://news.ycombinator.com/item?id=30458819

So I'm trying to figure out if setting up a new plant costs $100 million and takes a year to come online, and if there's no neon then we can't produce $500 million worth of chips a month. Why isn't some chipmaker or consortium fronting the $100 million and starting to build the plant now, as a cheap insurance policy to make sure they don't lose billions in revenue if the Ukraine production remains inaccessible for years?

> Why isn't some chipmaker or consortium fronting the $100 million and starting to build the plant now, as a cheap insurance policy to make sure they don't lose billions in revenue if the Ukraine production remains inaccessible for years?

Perhaps because there is a non-zero chance the disruption to the supply of neon will have disappeared long before the plant comes online, cheap supplies will be plentiful again, so they'll have "wasted" the $100 million ... and will have to explain to their investors why they took (with hindsight) such a crazy decision?

- "This high demand exists because neon is an essential, irreplaceable, and major component of argon-fluorine-neon excimer pulsed lasers."

Incidentally, that's the type of very high-power laser that was once considered indispensable for space-based missile defense satellites (like Polyus and the Star Wars concepts). To my understanding, this is the actual root cause of the current neon crisis: Soviet military overinvestment in the 1980's, creating a post-Cold War glut that priced out any new competition. Hence, the modern supply chain ended up relying on two countries that are now at war.

From this German government whitepaper about the noble gas industry:

- "Neon was regarded as a strategic resource in the former Soviet Union, because it was believed to be required for the intended production of laser weapons for missile and satellite defence purposes in the 1980s. Accordingly, all major air separation units in the Soviet Union were equipped with neon, but also krypton and xenon, enrichment facilities or, in some cases, purification plants (cf. Sections 5.4 and 5.5). The domestic Soviet supply of neon was extremely large but demand low."

- "Following the collapse of the Soviet Union, global crude neon production was approximately 500–600 million l/a (= 500,000–600,000 m3/a). It was dominated by far by large-scale air separation units associated with metallurgical combines in Russia and Ukraine. Simultaneously, demand was estimated at around 300 million l/a (cf. Section 4.2). In the years between 1990 and 2012, therefore, most crude neon was not purified, but released into the atmosphere, because there was no customer base."

https://www.deutsche-rohstoffagentur.de/DE/Gemeinsames/Produ... (chapter 5.2)

(By the way, this source links neon production to cryogenic oxygen separators at steel factories, whereas OP links it to Haber-Bosch plants (ammonia fertilizer). Anyone know which one is correct? Maybe both?)

Yeah, both : the nitrogen in ammonia comes from liquified air !

Polyus was a CO2 laser - do they need a lot of neon too?

> requiring a complete replacement roughly every two weeks, or complex and costly purification and recycling of the gaseous blende. It is a cost-benefit analysis for the $500 billion USD semiconductor industry.

Is the recycling process only available at lab scale currently due to the cost? How hard would it be to scale up? Or are chip companies betting on the war in Ukraine being over in a couple of months?

Seems that recycling is really well established/developed and is really just a minor inconvenience and a cost that the chip makers will get to once they can't proceed as usual.

https://www.photonicsonline.com/doc/how-one-light-source-man... https://asianometry.substack.com/p/neon-shortages-in-semicon...

I don't understand how the "purification" process would be substantially more difficult than distilling neon from air. If the waste blend has a higher percentage of neon than air, it should probably be cheaper/easier than buying new. I suspect that the real inconvenience is in the capture systems, another set of tubes in the chip fab needing servicing and occasional downtime.

Are they seriously venting fluorine?

The MSDS on Fluorine Excimer Laser Mix [0] lists fluorine at less than 1% concentration, and includes the note on reactivity: it produces some liquids that shouldn't persist in the air, these aren't CFCs. Guess it depends on the quantity but sounds like a rare event.

> Reactivity Hazards: Due to the presence of Fluorine, this gas mixture may react with water or moisture in the air to form hydrogen fluoride or hydrofluoric acid, plus small amounts of ozone, hydrogen peroxide and oxygen fluoride.

[0] https://lpg-umr6112.fr/lpg/fichiers/Intranet/FDS/FDS-Gaz/ArF...

Presumably the fluorine part gets scrubbed first?

This paper says they normally use KOH:


End result would be a solution of KF in water, which is benign.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact