How Dark Matter Physicists Score Deals on Liquid Xenon

How Dark Matter Physicists Score Deals on Liquid Xenon

Researchers assemble a prototype for their dark matter detector’s core, known as a time projection chamber.
Christopher Smith/SLAC National Accelerator Laboratory

If you want to build and run a $70 million dark matter detector, you're going to have a hefty shopping list. You'll need to buy hundreds of photomultiplier tubes, set up elaborate electronics, and pay graduate students, for starters. And 20 percent of your cash is going to go to just one thing: xenon gas. You'll need 200 steel bottles of the stuff, purified from the Earth’s atmosphere, at a price that can fluctuate wildly around $100,000 a bottle.

The physicists working on a project called LUX-ZEPLIN discovered this in 2015, when they were were developing their budget. The group is currently installing a giant dark matter detector underground in a former mine in South Dakota, where they hope to catch a glimpse of a hypothesized dark matter particle called a WIMP. And the xenon bills. Ouch. “No, we don’t get a break because we’re poor, starving scientists,” says physicist Murdock Gilchriese of the 250-member experiment.

Only a handful of companies produce xenon gas globally, by extracting it from the air at so-called air separation plants. Xenon makes up about 0.00001 percent of Earth’s atmosphere, a concentration comparable to a pot of soup seasoned with a single grain of salt. For these companies, xenon production is just a side gig, amounting to pocket change: Their more lucrative product is pure oxygen, which the steel industry buys to power their furnaces. “Nobody builds a plant just for xenon,” says Richard Betzendahl, a gas industry consultant and distributor. “There’s no economics in that.”

But it’s the lifeblood of dark matter experiments. Over the last decade, physicists have built increasingly gargantuan detectors filled with cooled, liquefied xenon to search for dark matter, a hypothesized substance that physicists think makes up 85 percent of the universe’s mass, according to decades of telescope observations. The detectors sense flashes of light that should appear when a dark matter particle bumps into a xenon nucleus. They’ve chosen to use xenon because it’s not reactive, which means less noise in their detector, and because it has a large nucleus, which makes it easier for dark matter to hit.

When LUX-ZEPLIN starts taking data in 2020, it will need 10 tons of xenon. They’ll have purchased xenon gas, purified it even further, and liquefied it for their detector. Other groups, like the XENON collaboration based in Italy, are planning even bigger detectors; the more xenon you use, the more likely you’ll catch a dark matter particle.


Several years ago, Gilchriese’s group did the math. The world produces about 40 tons of xenon a year. LUX-ZEPLIN needed about a quarter of the world’s yearly production.

Then, they went shopping.

They started by hiring Betzendahl as a consultant. “I didn’t know much about the xenon market,” says Gilchriese. Xenon, it turns out, has a variety of uses ranging from the mundane to the cosmic. Light bulb manufacturers put xenon gas inside incandescent bulbs to make the tungsten filament last longer. Space missions also use xenon as rocket fuel to launch satellites.

They also learned that they needed to time their bid just right. The price of xenon peaks and falls like a seismograph. “It can be 10 times what it was the previous year,” says Betzendahl.

That’s because in recent years, companies have developed, adopted, and abandoned new technologies that require large amounts of xenon. In 2008, Toshiba ordered a quarter of the world’s xenon supply for a new semiconductor technology, which caused the price of xenon to increase tenfold. “Then, they figured out how do it with a lot less xenon, and the price collapsed again,” says Betzendahl. Today, prices are rising, perhaps because Samsung has started buying xenon for a new etching technology.


So Gilchriese’s group, on Betzendahl’s recommendation, asked companies to deliver the xenon over several years. LUX-ZEPLIN gets its money from the government in batches: a few million the last three years. “Even if we had the money to buy it all at once, it was very clear that it would disrupt the market,” says Gilchriese. They also had to buy the xenon from multiple companies, largely based in the US and China. And because of Betzendahl’s advice, and maybe just luck, they managed to buy most of their xenon when prices dropped.

Gilchriese is done buying xenon, at least for a few years. But physicists are pondering the construction of more xenon-based dark matter experiments. The PandaX experiment in China is planning to build a detector containing four tons of liquid xenon. The XMASS collaboration in Japan aspires to a build a 20-ton detector. XENON has broached the possibility of a 50-ton detector. Gilchriese has asked one of his xenon dealers how that purchase might work. The verdict? “It’ll be really hard to get 50 tons of xenon,” he says.

They’d have to buy it in batches again, over a longer period of time, and hope they can buy it all at a fixed price. Even though physicists want more of it, it’s unlikely that companies are going to churn out significantly more xenon, says Betzendahl. In the three years or so it takes to plan and build a plant with xenon extraction capability, xenon prices could plummet. Lots of air separation plants don’t even bother extracting xenon and just let the gas escape back into the atmosphere.

More on Dark Matter

And xenon customers don’t have a lot of options. Some people in the xenon industry have toyed with the idea of extracting it from natural gas, where it occurs in higher concentration than regular air. But right now, air separation plants are literally the only way to harvest xenon, says Betzendahl.

In the early days of the LUX-ZEPLIN experiment, Gilchriese’s group actually considered building their own xenon production plant. Worried about the fluctuations in the xenon market, they wondered if they could just do it themselves. “Admittedly, it was a stab in the dark,” says Gilchriese.

They asked some engineers to estimate how much it would cost. “It just was not feasible,” says Gilchriese. “It was many tens of millions of dollars just to set up the plant. And you have to design it. Where would it go? And then you actually have to run it. You know what I mean? You’re certainly not going to compete with people who do this as an add-on to a much larger market.”

On the bright side, dark matter experiments don’t consume xenon—the 10 tons just stay in the detector, inert, in LUX-ZEPLIN’s detector for its five years of operation. “Someday they’ll have a lot of product to sell on the market,” says Betzendahl. They could recycle the xenon in another experiment. Or they could hoard it and wait for the price to spike again.