The Temporal Scan: Zeros as Dynamical Phase Transitions

From energy levels to instants

Chapter 31 left the Hilbert–Pólya conjecture standing as the best guess about why the critical line is a line: find a self-adjoint operator whose eigenvalues are the imaginary parts of the zeros, and the reality of those eigenvalues forces the zeros onto Re(s) = 1/2. A century of that search has produced deep structure — Berry–Keating semiclassics, the Connes trace formula, the Bender–Brody–Müller Hamiltonian — and no operator anyone can hold.

The move worth recording here is a change of variable rather than a change of ambition. Instead of asking which fixed spectrum contains the zeros, prepare a quantum state, quench it, and let it evolve. Track the return amplitude — the overlap of the evolved state with where it began, the Loschmidt amplitude — as a function of time. Its logarithm behaves like a free energy with time in the role usually played by inverse temperature, and where that function fails to be analytic the system is said to undergo a dynamical quantum phase transition. Heyl's framework for this is a decade old and well tested; the zeros of the Loschmidt amplitude in the complex-time plane are the Fisher zeros, and their crossings of the real time axis are what produce the observed kinks.

So the question becomes: can a system be built whose return amplitude is the zeta function along the critical line, so that its dynamical transitions occur precisely at t = 14.134…, 21.022…, 25.010…? The answer is yes, and it has been done.

What was actually built

Wei, Zhai, Lu, Yang, Gao, Wei, Song, Nori, Xin and colleagues — a BAQIS/Tsinghua/SUSTech/RIKEN collaboration — published the realisation in 2025 and it appeared in Nature Communications in 2026 under the title ‘The Riemann Hypothesis manifested in dynamical quantum phase transitions’, with an accompanying preprint treating the mapping through fractal manifolds. The construction encodes the Dirichlet-series structure of ζ into the spectrum and initial-state weights of a controllable quantum system, so that the Loschmidt amplitude reproduces ζ(1/2 + it) as the system evolves. The non-trivial zeros then appear as genuine non-analyticities in the dynamical free energy — measurable kinks, in a laboratory, at the arithmetic times.

This is a serious and beautiful piece of work, and its value is precise: it converts a statement about the location of zeros into a statement about the criticality of a physical evolution, and it makes the zeros experimentally addressable in a system a person can build. If a zero were to sit off the critical line, the corresponding signature in the constructed system would be a transition that fails to reach the real-time axis — a missed kink. That is a physical restatement of the hypothesis, and restatements of this kind have historically been where progress comes from.

What it is not is a discovery that quantum fields in the wild transition at zeta times. The zeta function is the input. The experimenter chooses couplings and weights so that ζ comes out; the zeros are then read off an amplitude that was engineered to contain them. Calling this a detection would be like calling a slide rule a discovery of logarithms.

The scan image, kept and bounded

With that stated, the image of a temporal scan is worth keeping, because it is the correct picture of the mathematics. Evaluating ζ up the critical line is a sweep in one real parameter. Under the dynamical reading that parameter is time, the sweep is an evolution, and the zeros are the instants at which the evolution goes critical. The volume's recursion/iteration vocabulary fits without strain: the multiplicative structure of the primes is the recursive input fixing the spectrum; the forward march of t is the iteration; the zeros are the terminating values where the iteration produces something that persists.

The stability argument also survives in its proper form. Because the zeros lie on a line, the sequence of transition times inherits the rigid spacing statistics of Chapter 31 — GUE repulsion, no clustering, no gaps of arbitrary size. A drifting zero would break the regularity of the transition sequence in the constructed system. That is a statement about the constructed system, and it is true of it.

What the chapter will not do is promote the localisation language. ‘Matter freezes out of the fluid at a zero’ is not what the experiment shows. A dynamical quantum phase transition is a non-analyticity in a return amplitude; it is not the appearance of a particle, and no mass, charge, or lifetime is produced at a Fisher-zero crossing. The honest version of the sentence is that the zeros mark where a prepared quantum evolution changes character, and that the model of Chapters 25–31 reads localisation into that change without having earned the operator that would license it.

What would raise this from restatement to physics

Three things, none of which is in hand. First, a system in which ζ is an output rather than an input — where the arithmetic emerges from independently motivated dynamics instead of being encoded in the couplings. Second, a dimensional bridge: t in ζ is a pure number, and until there is a stated scale converting it to seconds, ‘t = 14.134’ is not a time. Third, a link from the non-analyticity to an order parameter with physical content, so that ‘the field's characteristics change’ names a measured observable rather than a change in a mathematical function.

The first is the one to watch. The second is a bookkeeping demand that any physical reading must satisfy and that the model has never satisfied. The third is where the word ‘matter’ in the mental model would either acquire content or be withdrawn.