Future paths

Where matter could go next

The glass on your phone, the magnet on the fridge, the frame of your bike: each began as a strange result on a lab bench. Seven such results sit on benches right now. None is promised; each is being worked at, and any one of them would change the shape of an ordinary day.

7paths under way, from robot chemists to walls built to face a small star
41compounds a robot line in California cooked for the first time, across 17 days of continuous running
1open question holding up each of the seven paths; the whole field runs on them

Seven paths

Every path below has two sides: what is on the bench now, and what would land at your place if it worked. Flip any card to see the other side.

๐Ÿฆพ

Self-driving labs

A robot arm weighs the powders, a furnace cooks them, an instrument reads what formed, and software chooses the next recipe while the town sleeps. The loop of propose, cook and test runs nights, weekends and public holidays. One such line in California reported 41 new compounds, blends of elements locked into new solids, across 17 days of continuous running.

Most objects in your house waited decades between somebody's good idea and the shelf. If the propose-cook-test loop keeps speeding up, that wait could shrink: a fridge coating that sheds frost, a cheaper solar tile, a longer-lived hot water heater element could each arrive years sooner than the old pace would have carried them.

Open questionCan software tell a genuinely new material from a burnt mess in the crucible, without a chemist leaning in for a look?

๐Ÿ”ฉ

High-entropy alloys

Nearly every metal object you own is one main metal with a pinch of helpers: steel is iron with a whisper of carbon. High-entropy alloys turn that recipe upside down. Five or more metals blend in roughly equal shares, like a fruitcake where every slice holds the same mix. Some blends grow stronger, not weaker, in deep cold: one chromium, cobalt and nickel mix posted toughness readings, its knack for bending rather than snapping, among the highest on record, in tests at about minus 253 degrees Celsius. That is 20 kelvin, the scale that counts up from absolute zero, so twenty degrees above the bottom of the temperature scale.

Blades that hold their edge through a year of Sunday roasts. A tow bar that stays tough on the frostiest morning of a high-country winter. Turbine parts that run hotter and turn more fuel into power instead of waste heat, which would show up quietly on the power bill. Cold-tough steels would also suit the trucks and freezers that carry food and medicine to your shops.

Open questionMillions of five-metal recipes are possible; which handful deserve the furnace time?

๐Ÿ”‹

Solid-state batteries

The battery in your phone ferries charged atoms through a flammable liquid: the electrolyte, the syrup the charge swims through. Solid-state cells swap that liquid for a thin ceramic or plastic that stays put and still lets the charged atoms slide. Small cells work in labs today; the climb is making them by the million, cheaply, in sheets big enough for a car.

A phone that would shrug off a hot dashboard. An electric ute that could take on charge in about the time a counter coffee takes, and hold more of it, so the commute and the weekend trip fit inside one battery. A home battery that would sit in the garage with one less flammable thing inside it.

Open questionBatteries breathe, swelling and shrinking with each charge; can a solid keep snug contact through thousands of those breaths?

๐Ÿฅž

2D stacking and twistronics

The graphite in a pencil is a deck of carbon sheets, each a single atom thick. One sheet on its own is graphene, so thin that about three million of them stack to a millimetre. Lay one sheet on another and turn the top one by about a degree, and the pair behaves like a brand-new material: within a degree or two of absolute zero, where every Celsius reading has bunched up near minus 273 and the count from the bottom of the scale is the one worth reading, one twisted pair carries current with no measurable loss. Choosing the sheets and the angle of the turn is a craft of its own, called twistronics.

Materials tuned by geometry rather than by chemistry could be dialled in like a radio station: a window film that switches from clear to shady on a scorcher, sensors fine enough to read a heartbeat through a shirt, chips that run cooler in the laptop on your knees.

Open questionThe effects live in flakes smaller than this full stop; whether anyone could make them by the metre is wide open.

โšก

Room-temperature superconductors

A superconductor carries electric current with no measurable loss. Many materials manage it today, but only when chilled colder than any winter on Earth. Squeeze certain hydrogen-rich compounds to more than a million times the push of the air around you, pressures like the deep inside of the planet, and the trick survives up to about minus 23 degrees Celsius (250 kelvin), the chill of a shop ice-cream freezer. The prize would be a wire that does it on a bench: no chill, no squeeze. None is confirmed at everyday pressure so far. Claims of exactly that have appeared, and each has gone onto other benches, where the resistance drop traced back to something else in the sample. Those traces are published, which is why each round of checking runs faster than the last.

Roughly one part in twenty of the electricity sent down the wires arrives as warm cable instead of a boiled kettle. Wires with no loss would close that gap on the power bill, and would open doors well past it: trains that float on magnetic fields for a smoother commute, medical scanners cheap enough for country hospitals, motors the size of a mango.

Open questionIs there a material family that keeps the trick at everyday pressure and temperature, or is there a ceiling built into the way electrons pair? Nobody has written that ceiling down, and nobody has made the wire. Both halves of the question are still open.

โญ

Materials for fusion

A fusion machine holds a working star. Its fuel is plasma, a gas so hot its atoms have come apart, at more than 100 million degrees: over six times the heat at the centre of the Sun. That high up, Celsius and kelvin read the same number, since the 273 between the two scales vanishes into the rounding. Magnetic fields hold the plasma away from the walls, yet the walls still cop the glare plus a hail of neutrons, the uncharged pieces that fly out of the reaction. Tungsten, the metal with the highest melting point at 3,422 degrees Celsius (3,695 kelvin), about three times the heat of fresh lava, lines the hottest parts of today's test machines. Steels that stay sound after years of that hail are still being worked out.

If the walls hold, fusion stations could one day feed the grid: steady power for the fridge, the kettle and the air-con, drawn from a fuel found in seawater, running through rain, hail and still nights. That is a could, not a will; the walls are a fair share of the distance left.

Open questionNo test rig yet matches decades of neutron hail; which steel would still be sound at year twenty?

๐Ÿ’ 

Discovery at GNoME scale

Before anything gets cooked, software that has studied the known crystals can sketch which new recipes would likely hold together. One effort, called GNoME, predicted about 2.2 million new crystals, roughly one for every person in Perth, and flagged about 380,000 of them as likely stable: give each crystal a seat and they would fill the MCG (the Melbourne Cricket Ground, about 100,000 seats) close to four times over. Hundreds from the list have since been made in labs and held together as the predictions said they would.

A shortlist that long would keep the robot labs fed for decades. Somewhere in it could sit the solid for the next battery in your ute, the blend for a tougher kitchen knife, the wall for the small star. The shortlist exists now; the queue for the furnace is the story of the coming years.

Open questionLikely stable on paper is not the same as makeable in a crucible; how much of the list can be coaxed into existence?


Where to look tomorrow

Neither of these waits on a lab coat.

๐Ÿ—‚๏ธ

Browse the Materials Project

A free public database holding well over a hundred thousand materials, each with its properties laid out: more entries than an MCG grand final crowd, one material per seat. Free to browse, and the same well many of the labs above drink from.

๐ŸงŠ

Grow a crystal on the windowsill

Crystal growing kits are real and sit on school supply and toy shop shelves. Sugar on a string in a jar of warm syrup works too. In about a week a faceted crystal forms, its atoms filing into the same ordered ranks the labs above spend careers arranging.

Sources

  • A-Lab robot chemistry line at Lawrence Berkeley National Laboratory: 41 new compounds in 17 days, Nature, 2023.
  • GNoME crystal predictions, Google DeepMind, Nature, 2023: about 2.2 million candidates, about 380,000 flagged stable, 736 since made in labs.
  • Chromium cobalt nickel alloy toughness measured at 20 kelvin, about minus 253 degrees Celsius, Science, 2022.
  • Superconductivity in twisted bilayer graphene, Nature, 2018.
  • Superconductivity in lanthanum hydride near minus 23 degrees Celsius (250 kelvin) under about 170 gigapascals, Nature, 2019.
  • The Materials Project, materialsproject.org, free to browse.