Papers on what gravity does and does not do to quantum matter: where a Diósi–Penrose decoherence rate cannot come from, what classical–quantum gravity would need to survive the X-ray data, when the Diósi–Penrose model can entangle two masses, and how matter absorbs a gravitational wave.
Foundations · Sum ruleNovember 17, 2026Working paper
Graviton Absorption and the Slope of the Gravitational Form Factor
How strongly can a piece of matter absorb a gravitational wave? For a weak plane wave and any matter with a conserved stress tensor we show that, to first order in Newton's constant and without a multipole expansion, the absorption cross-section is the spectral function of one operator: the quadrupole moment, across the direction of travel, of the stress tensor integrated along the null rays of the wave, the averaged null energy. Its moment ∫σ dω/ω² gives the energy left by a wave with a sudden kink and is the expectation value of a double commutator. For slow matter in a body small against the wavelength the double commutator is the moment of inertia about the direction of travel, which gives ∫σ dω/ω² = 2π²GI/c³, the gravitational counterpart of the Thomas–Reiche–Kuhn rule and in content the energy-weighted quadrupole sum rule of nuclear physics; we check it on hydrogen and on a bar detector.
For relativistic matter in canonical light-front quantization, with interactions that are local across the ray, it is the second transverse moment of the averaged null energy, and for a neutral particle of mass m and spin 0 or 1/2 at rest, ∫σ dω/ω² = 8π²Gm A′(0) in units ħ = c = 1, with A(t) the gravitational form factor; the D-term and the trace form factors do not enter. We verify this at second order in the coupling for a scalar and a Yukawa theory by computing both sides with covariant Feynman rules: they agree to the accuracy of the quadrature. At that order the interaction enters only through the particle's wavefunction, so the test confirms the chain of statements and not the algebra behind its extension. Beyond that order the evaluation rests on the canonical algebra of light-ray operators at coincident transverse points, which is known not to be universal; we state the rule with its explicit correction term and leave its status in renormalized perturbation theory open. Draft; working paper; not peer-reviewed.
@article{Sperzel2026GravitonAbsorption,
author = {Sperzel, Marc},
title = {Graviton Absorption and the Slope of the Gravitational Form Factor},
year = {2026},
note = {Quantum-Geometric Correspondence Series},
url = {https://www.marcsperzel.com/papers/html/graviton-absorption-sum-rule.html}
}
In classical–quantum gravity the metric stays classical, and complete positivity ties the decoherence of matter to the diffusion of the metric. The covariant theory of Oppenheim and collaborators is viable by its authors' assessment, which does not evaluate the matter noise at keV frequencies. That noise conflicts with XENONnT's search for spontaneous X-rays, by two arguments that fail under different escapes.
First, the covariant diffusion makes the conformal mode dynamical; near the light cone slow matter feels it through its rest mass, while Maxwell's equations do not. For translation-invariant kernels and every diffusion weight that is Lorentz invariant in the scalar sector, whatever its regulator or running in p², the trade-off forces matter noise s ≥ (2/3)|f|/p² on a conserved near-null source, in units of the Diósi–Penrose kernel. Where Newton's law is tested, |f| = 1 and ∫s dp² ≥ 9.2, against a limit of 7.7×10⁻⁸ from our likelihood fit to the published XENONnT spectrum: an exclusion by ≥ 1.1×10⁸ (median expected 6.6×10⁷). A frame in the scalar weight evades it.
Second, general relativity's response to a conserved near-null source has a light-cone pole, which is purely tensor. The resulting bound holds for every positive diffusion weight, with or without a frame. It assumes general relativity's tensor propagator at keV momenta, but no transfer of static tests; an atom's Coulomb field gives the tensor source an order-one share. The margin, 1.7×10⁹ (median expected 9.8×10⁸), is an estimate, not a bound.
At linear order, a classical theory that survives both needs a frame or a non-perturbative mass above (0.34 µm)⁻¹ in the scalar sector, and a tensor sector off the light cone, or with a form factor, at keV momenta, or a coupling blind to the atom's stress; or a frame carried by the matter coupling, or possibly by the state of an extra field (open); or noise outside the Gaussian class, which for a Poisson record of jumps must deliver its keV emission in multi-photon bursts. Draft; working paper; not peer-reviewed.
Double-Pulsar Drag Bounds Dissipative Sources of Diósi–Penrose Decoherence
In the proposal of Diósi and Penrose a mass in a superposition of two positions loses coherence at the rate Γ = C·E_G/ħ, first order in Newton's constant. We ask what could produce this rate for a static superposition of equal masses. A bath in its ground state cannot: its noise equals its absorption, which vanishes at zero frequency when the response is continuous there. For the free linearized gravitational field and a 1 µg body the loss of coherence stays below 10⁻² at all times, for separations up to a millimetre, in a scalar proxy normalized to E_G (it tends to a few times 10⁻³), and it is reversible. These results are standard, at second cumulant. A bath that is not in a ground state can supply the rate, but if it couples to the mass density through a Hamiltonian, the noise that decoheres a body also takes momentum from it. We show that stationary, white, passive noise of Diósi–Penrose form drags a moving body at a rate of at least (2π/3)(CħG/k_B T_eff) times its density, with T_eff the fluctuation–dissipation temperature. If the noise keeps this form and one temperature out to the kilometre wavelengths that a neutron star resolves, as it does by construction in the linear-friction dissipative Diósi–Penrose model, the orbital decay of the double pulsar PSR J0737−3039A/B requires T_eff/C > 3.3×10¹⁶ K, up to strong-field corrections of order one. This is more than 5×10²⁷ above the laboratory bound on that model and excludes the proposed T_β ≈ 3 K by 10¹⁶. Baths that differ between kilometre and laboratory scales escape.
Otherwise what survives is noise without back-action: Diósi's original model with Gaussian mass-density smearing R₀. Published data allow R₀ > 4.5 Å·C^(1/3) with no upper bound and do not fix C. For a uniform sphere of radius a ≫ R₀ the rate is independent of R₀ to 1.7 R₀/a at every separation d, so one visibility measurement, with other decoherence channels subtracted, determines C; for 1 µg of silica τ = 65 ps/C at d = 1 mm and 0.35 µs/C at d = 1 µm. Working paper; not peer-reviewed.
Entanglement from the Diósi–Penrose Model: An Exact Criterion, and Why It Stays Small at Any Distance
Trillo and Navascués showed that the Diósi–Penrose model, in the Tilloy–Diósi form with the same smearing length σ in force and noise, entangles two masses in superposition when they are closer than about 2.2 σ, and asked whether gravitationally induced entanglement can be observed beyond that distance. We answer for two rigid bodies whose branches are position eigenstates on finite sets, without kinetic term, for noise strength C. One kernel, the Coulomb energy of the smeared bodies, sets both the Newtonian coupling and the correlated noise, and the question becomes a single geometric comparison: a product state becomes entangled if and only if the largest canonical correlation between neutral charge patterns on the two branch sets exceeds C/√(1+C²). Below this threshold the evolution is a separable operation at all times.
The published critical distances are two-branch cases of the criterion: for parallel two-branch superpositions the supremum is 2.21 σ at C = 1, and a straddling geometry reaches 2.74 σ. With many branches there is no critical distance. What limits the effect is the model's own dephasing, which sits in the same quadratic form. For bodies of effective size σ_eff (σ for point masses) separated by a slab of width d ≫ σ, the entangling rate at C = 1 is at most 0.12 σ_eff/d times the single-body dephasing rate, to leading order in σ/d. For branches more than σ_eff apart the entanglement is therefore transient and, if well separated, a fraction of order σ_eff/d of what unitary Newtonian gravity gives; below σ it can last for days. Entanglement of the size unitary gravity predicts still excludes the model; the inference from any detected entanglement beyond a fixed multiple of σ does not hold. Draft; working paper; not peer-reviewed.
@article{Sperzel2026DPEntanglement,
author = {Sperzel, Marc},
title = {Entanglement from the Diósi–Penrose Model: An Exact Criterion, and Why It Stays Small at Any Distance},
year = {2026},
note = {Quantum-Geometric Correspondence Series},
url = {https://www.marcsperzel.com/papers/html/dp-entanglement-criterion.html}
}
Lecture Notes
Two self-contained courses on the theory behind the papers: a ground-up volume that assumes only undergraduate physics and builds everything else from scratch, and a compressed six-lecture review for readers who already carry the graduate toolkit.
Foundations Track — a fully self-contained course in fifteen chaptersOct 2026 (rewritten) · 270 pp
The Quantum-Geometric Correspondence from the Ground Up
Assumes only undergraduate quantum mechanics, special relativity, and calculus. Everything else is built before it is used.
A complete, ground-up course on the theory behind the papers. Part I states the problem: the two best theories in physics contradict each other on where a superposed mass gravitates. Part II builds the full toolkit from scratch — density matrices and decoherence, general relativity, quantum fields and the Unruh effect, black-hole thermodynamics, operator algebras, and modular theory. Part III states the correspondence, lays down its three axioms, derives Einstein’s equations from entanglement equilibrium, and shows with no skipped steps where a gravitational-decoherence rate cannot come from. Part IV asks what an observed rate would mean: the chain to a preferred frame, the two branches (a classical record or quantum gravity), and the experiments that decide between them. Every objection a sharp reader would raise is asked and answered on the spot in numbered question boxes; every claim carries an honest status label (proven / derived / conjectured / assumed); every number is recomputed from G, ħ, c and machine-checked against the repository’s verification harness.
Review Track — the research corpus in 41 dense pagesOct 2026 (rewritten) · 41 pp
The Quantum-Geometric Correspondence: A Six-Lecture Course
For readers who already carry graduate quantum field theory, general relativity, and operator algebras.
The same theory compressed into six lectures in the register of a mathematical-physics monograph: definitions, axioms, theorems and derivations, numbered equations, exact status labels. Covers the kinematic foundations, the three axioms of the Modular Diamond formulation, the no-go theorems and the rate postulate, the operator-algebraic frontier, cosmology and the axioms, and the phenomenology-and-status ledger. Each lecture ends with a hands-on layer of worked exercises keyed to the repository’s verification scripts.