Our knowledge around them is quite new. First theorized only in 2012 and first "experimentally realized" in 2016. The novelty of it all does evoke a kind of wierdness. A decade later and we're using them in quantum computers. The future is now and it's sci-fi, man.

The Wikipedia article (wikipedia.org) provides a neat overview of the "what" without waxing too technical. It fails to satisfy my nagging need to answer "how?!" though.

This article (technologyreview.com) provides a decent answer for how time crystals are possible in a lab.

Their quantum system is a line of ytterbium ions with spins that interact with each other. That interaction leads to a special kind of behavior. ... One of the key properties of these ions is their magnetization or spin, which can be flipped up or down using a laser. Flipping the spin of one ion causes the next to flip, and so on. These spin interactions then oscillate at a rate that depends on how regularly the laser flips the original spin. In other words, the driving frequency determines the rate of oscillation.

But when Monroe and co measured this, they found another effect. These guys discovered that after allowing the system to evolve, the interactions occurred at a rate that was twice the original period. Since there is no driving force with that period, the only explanation is that the time symmetry must have been broken, thereby allowing these longer periods. In other words, Monroe and co had created a time crystal.

In trying to understand this I ended up reading and digesting the following: Physics: Time Crystal (handwiki.org) Spontaneous symmetry breaking (wikipedia.org) Symmetry (physics) (wikipedia.org) In Search of Time Crystals (physicsworld.com)

Quoting from that last article:

To understand time crystals, let’s remind ourselves about ordinary crystals. Diamond, say, breaks spatial symmetry because not every location is equivalent. Some locations have carbons atoms; others don’t. If you shift, or “translate”, the diamond lattice by some arbitrary amount, it won’t superimpose on the original lattice; the crystal structure has broken the translational symmetry of uniform space. But if you shift the lattice by some integer multiple of the spacing between atoms, it does superimpose, which means that the broken translational symmetry is periodic.

A "time crystal" breaks translational symmetry in time rather than space. This creates a kind of clock analogous to chemical oscillators (wikipedia.org). To keep a chemical oscillator going though one must continue to add reagents because the system is burning energy. Theoretically time crystals are stable in perpituity at equilibrium. Their lowest energy state includes motion.

The time crystals discussed in the Physics World piece and elsewhere, so far as I can find, are all "discrete" time crystals. These are driven by an external force. So they aren't in equilibrium... But these are still curious for two reasons. First, as mentioned, is that changing the driving frequency does not change the frequency of oscillation. The second is that discrete time crystals don't seem to be absorbing the energy imparted to the system by the driver (a laser, microwaves, etc). They're not heating up. It doesn't seem like we're quite sure why, either. More mysteries for humans to practice science around!