Characteristics of mysterious time crystals have been found in the most unexpected place you'd ever think to look - a compound found in fertiliser and those crystal-growing kits you can buy for kids.
That compound is monoammonium phosphate (MAP), and the physicists from Yale who made the discovery are now scratching their heads, because this raises questions about how time crystals even form in the first place.
In normal crystals, the atoms are arranged in a fixed grid structure, like the atomic lattice of a diamond or quartz crystal. These repeating lattices can differ in configuration, but they don't move around very much - they only repeat spatially, but not in time.
To the naked eye, they look like ordinary crystals. But their atoms are actually oscillating - spinning first in one direction, and then the other, when exposed to an electromagnetic pulse that flips the spin.
Even when the pulse is irregular, the oscillations - what the researchers refer to as the time crystal "ticking" - is locked to a particular, and very regular, frequency.
Because they're so new, discrete time crystals (DTCs) haven't been observed very often, and only once before in a solid crystal, when Harvard physicists created a time crystal from a nitrogen-vacancy diamond.
The first experiment and the University of Maryland in 2016 demonstrated time crystal behaviour in a line of ytterbium atoms.
The new time crystal found at Yale, although inspired by these experiments, is different yet again from those first two discoveries.
"We decided to try searching for the DTC signature ourselves," said physicist Sean Barrett, senior author on two new papers.
"My student Jared Rovny had grown monoammonium phosphate (MAP) crystals for a completely different experiment, so we happened to have one in our lab."
Previously, it was thought that time crystal signatures could only occur within a more disordered environment.
However, after subjecting the monoammonium phosphate crystals to nuclear magnetic resonance, the team found clear time crystal signatures - inside a highly ordered spatial crystal.
"Our crystal measurements looked quite striking right off the bat," Barrett said. "Our work suggests that the signature of a DTC could be found, in principle, by looking in a children's crystal growing kit."
Time crystals have great potential for practical applications. They could be used to improve our current atomic clock technology - complex timepieces that keep the most accurate time that we can possibly achieve.
They could also improve technology such as gyroscopes, and systems that rely on atomic clocks, such as GPS - and they could aid in quantum entanglement experiments. Even DARPA is pouring resources into researching time crystals - although remaining tight-lipped about why.
So the possibility that they can occur in such a common and seemingly normal crystal is exciting.
But the work also presents a puzzle. If time crystals can occur within ordered spatial arrangements in ordinary crystals, physicists will have to figure out how this occurs - and why a greater number of ordinary crystals don't exhibit time crystal signatures.
"It's too early to tell what the resolution will be for the current theory of discrete time crystals, but people will be working on this question for at least the next few years," Barrett said.