In 1867, James Clerk Maxwell imagined a tiny creature sitting at a door between two chambers of gas. The creature could see individual molecules and selectively open the door, letting fast molecules through one way and slow molecules the other. Without doing any work, the demon would make one chamber hot and the other cold — violating the second law of thermodynamics. Maxwell called it a "finite being." Lord Kelvin named it a demon.
Eighty-nine years later, in 1956, the physicist David Pines borrowed the name. He was studying what happens inside metals when electrons move collectively — not as individual particles bouncing around, but as coordinated waves of charge density called plasmons. Pines and David Bohm had proposed plasmons in 1952: quantum units of collective electron oscillation, the same way phonons are quantum units of vibration in a crystal lattice.
But Pines noticed something strange could happen in metals with more than one type of electron band. In such metals, electrons in different bands can oscillate together. If both bands oscillate in phase — peaks aligning with peaks — you get the ordinary plasmon, a massive excitation at high energy. But if the two bands oscillate out of phase — the peaks of one coinciding with the valleys of the other — something uncanny emerges. The charges cancel. The mass vanishes. The result is a collective excitation that is electrically neutral, massless, and invisible to light.
Pines called it a demon.
Consider what it means for a thing to be invisible in this way. A plasmon is a ripple in the electron sea, and because it carries net charge, it interacts with electromagnetic radiation — you can detect it with light. Pines' demon carries no net charge. The two bands of electrons are sloshing back and forth in opposite directions, and at every point their charges exactly screen each other. No net charge fluctuation, no coupling to photons. The demon is, in the most literal sense, dark.
It is also massless. Ordinary plasmons have a minimum energy even at zero momentum — a "gap" that arises from the long-range Coulomb interaction between charges. But the demon, being neutral, feels no such pull. Its dispersion relation is acoustic: energy proportional to momentum, like sound. Except the demon travels at roughly a hundred times the speed of sound in the crystal. A massless, chargeless wave screaming through a lattice at velocities between sound and light.
For sixty-seven years, this particle existed only in equations.
The problem with finding the demon was exactly its defining property: invisibility. Conventional spectroscopy uses photons, and the demon does not talk to photons. You cannot shine light on a metal and watch for the demon's signature. You need a probe that couples to density fluctuations regardless of charge — and in 2023, a team at the University of Illinois found one.
Peter Abbamonte, Ali Husain, and their collaborators used momentum-resolved electron energy-loss spectroscopy. They fired electrons into a crystal of strontium ruthenate (Sr2RuO4) — a layered material with multiple electron bands — and measured how much energy the electrons lost at each momentum. At high energies, around 1.2 electron volts, they saw the conventional plasmon: the in-phase oscillation, massive and visible. But at low energies, below 8 millielectron volts at zero momentum, they found something else. An acoustic mode. A ghost in the spectrum.
To confirm it was truly the demon, they studied how the spectral intensity varied with momentum. A charged excitation's intensity falls as the inverse fifth power of momentum. A neutral one falls as roughly the inverse square. The measured exponent was 1.83 — unmistakably neutral. Pines' demon, detected at last.
Why does this matter beyond the elegance of confirming a six-decade prediction?
Because of superconductivity. In a superconductor, electrons pair up into Cooper pairs — bound states that flow without resistance. The standard BCS theory says phonons mediate this pairing: one electron distorts the lattice, and another electron falls into the distortion. But BCS theory struggles with high-temperature superconductors, materials that superconduct at temperatures far above what phonon-mediated pairing can explain. Something else must be binding the electrons together.
Pines suspected his demon. A massless boson that exists at any temperature, that lives inside multiband metals, that couples to electrons but not to light — it is precisely the kind of mediator that could enable pairing at temperatures the lattice alone cannot sustain. Magnesium diboride, iron-based superconductors, possibly even the cuprates — all are multiband metals where demons might lurk.
The discovery in strontium ruthenate is proof of existence, not proof of mechanism. But it changes the landscape. We know now that demons are real, that they can be detected, that the theory is not merely elegant mathematics. The next question is whether they participate in pairing.
What strikes me most is the naming. Maxwell's demon was an impossibility — Szilard and Landauer eventually showed that the act of observing molecules generates enough entropy to save the second law. The demon cannot exist because knowing is not free. Pines' demon is the inverse: a thing that cannot be known by the usual means. Maxwell's demon fails because observation has a cost. Pines' demon hides because observation has a limitation.
Two demons, mirrored. One is defeated by information theory. The other is shielded by it.
There is something satisfying about a sixty-seven-year wait. Not because patience is virtuous, but because it demonstrates that a prediction can be precise enough to survive its predictor. David Pines died in 2018. Five years later, his demon was caught in a crystal in Urbana-Champaign — the same university where he spent most of his career. The mathematics outlived the mathematician, and the crystal confirmed what the equations already knew.
The demon is neutral, massless, and invisible to light. It exists between the peaks and valleys of two electron bands, in the exact place where their charges cancel. It is the ghost of a difference.