Explainer graphic: how a pillar template turns a layered eutectic into ordered patterns, with 50% and 99% order figures and four questions to ask
Same ingredients, different insides: a forest of tiny pillars decides the pattern a cooling material settles into.

Scatter salt on an icy road and you are doing materials science. Salt and water form what chemists call a eutectic: a mixture that melts or freezes at a lower temperature than either ingredient alone. A 20% salt solution freezes at about -16°C instead of 0°C, as the American Ceramic Society explains. Solder is another everyday eutectic.

The short answer: When a eutectic liquid cools, its ingredients sort themselves into a pattern without any help. That is self-assembly. Templated self-assembly means cooling the liquid inside a tiny scaffold, so the scaffold steers which pattern forms. In 2020 a US team showed this works for hard, inorganic materials. In 2024 they made the patterns far more regular. It is still lab research, not a product.

What self-assembly does on its own

Left alone, many eutectics separate into thin alternating layers as they solidify, a bit like a very small lasagne. Nobody places the layers. The atoms shuffle sideways ahead of the freezing front and the stripes simply appear.

That is cheap and fast, which is why engineers like it. Eutectics already turn up in solder alloys and in the turbine blades of jet engines. The catch is control. You get the pattern nature hands you, plus whatever flaws come with it. Paul Braun of the University of Illinois put it down to the "uncontrolled nature" of the process in the university's 2020 announcement.

What the pillars changed

Braun's group, working with Katsuyo Thornton's team at the University of Michigan, melted a mix of silver chloride and potassium chloride. On its own it forms plain layers. They poured it into a template: nickel pillars in a hexagonal grid, each 500 to 620 nanometres wide and 4 to 6 micrometres tall, according to MRS Bulletin.

The pillars got in the way of the usual sideways shuffle. Instead of stripes, the material settled into square, triangular and honeycomb arrangements. Outside the pillars, the ordinary layers carried on as if nothing had happened.

The clever part is that the result was not a copy of the template. First author Ashish Kulkarni described spoke-like shapes with three, four, five or six arms, and the team could switch between them by changing how fast the melt solidified. Same template, same ingredients, different insides. He tells the story in a Springer Nature post. The work appeared in Nature in January 2020.

The 2024 fix: stop the pillars carrying heat

The first patterns were pretty but not perfectly regular over long distances. The team later worked out why. The pillars were blocking atoms as intended, but they were also carrying heat, which bent the freezing front out of shape.

So they rebuilt the pillars from porous silicon, described as essentially a silicon foam, which conducts heat at least 100 times worse than solid crystalline silicon. The difference was large, as the Illinois Materials Research Laboratory reported when the study came out in Advanced Materials in February 2024.

How the template changes the result
SetupWhat formsHow regular
No templateAlternating layersWhatever nature gives
Nickel pillars (2020)Square, triangular and honeycomb patternsFewer defects, team reported
Crystalline silicon pillars (2024)Intended pattern, patchyAbout 50% of unit cells
Porous silicon pillars (2024)Intended patternAbout 99% of unit cells

Why the inside of a material matters

Two blocks can contain identical ingredients and still behave differently, because behaviour depends on arrangement. Think of two buildings with the same floor area but different corridors. How easily you cross them depends on the layout, not the bricks.

Here the patterns have features about the size of a wavelength of visible light. The salts used are transparent, so Braun pointed to optical parts such as photonic crystals as a likely first direction. Other combinations might one day guide heat as well as light.

Mind the word "might". A new pattern proves the team can build a structure. It does not prove the structure bends light or carries heat better. That needs its own measurement.

How to read the next materials headline

  1. Ask what was compared. Same ingredients and cooling, with and without the template?
  2. Look past the pretty image. How large was the area, and how often did the pattern fail? The 50% versus 99% figures answer exactly that.
  3. Separate structure from performance. "We made a honeycomb" is not "it works better".
  4. Count the extra steps. Someone has to make the template, and that belongs in any cost claim.

Frequently asked questions

Is templated self-assembly used in products yet?

Not for these materials. In 2020 Braun said, "We're still a long way from real application." The 2024 study improved the order of the patterns in the lab. It did not announce a product. Unlike silicon-carbon phone batteries, this one has not made the trip from lab to pocket.

Does it only work with those two salts?

The salts were picked because they are well studied and transparent. Kulkarni told MRS Bulletin that simulations suggest the principle should apply to any eutectic system, and the team said it was looking at magnetic alloys next. That is a prediction from modelling, so treat it as promising rather than proven.

So the next time a headline says scientists "control a material from the inside", you know the trick. A liquid, a scaffold and a carefully managed cool-down. Then ask the boring question: did anyone measure what the new pattern actually does?