Does quantum mechanics do functional work in the brain?

1Background: the quantum-mind hypotheses

That the mind might be quantum-mechanical at some functional level is an old and persistent idea. It must be separated immediately from a true but unrelated fact: the brain, like all matter, is made of quantum constituents. The substantive claim — the one this report assesses — is stronger and more specific: that coherent quantum effects (superposition, entanglement, or quantum computation) do functional work in cognition or consciousness that a classical description cannot capture. Two modern hypotheses make that claim with enough mechanistic detail to be tested.

1.1Orchestrated objective reduction (Orch-OR)

The best-developed version is the Penrose–Hameroff Orch-OR program.123 Penrose argued from the Gödel incompleteness theorems that human mathematical insight is non-algorithmic, and proposed that a new physics — an objective reduction (OR) of the wave function driven by gravity — underlies it.1 Hameroff supplied the biological substrate: the microtubules of the neuronal cytoskeleton, whose tubulin subunits he proposed could exist in superposition and perform quantum computation, “orchestrated” by the cell and terminated by Penrose’s gravitational OR.23

The proposed collapse time follows from the gravitational self-energy E_G of the mass distribution that is in superposition, via τ ≈ ℏ / E_G.13 For the theory to be relevant to cognition, that collapse must occur on the timescale of a conscious “moment” — tied to 40 Hz gamma-band activity, i.e. about 25 ms, with ~100 ms (10 Hz) as a looser bound.3 This 25 ms requirement is the load-bearing number for the whole program, and the target of the standard objection (§2).

1.2The Fisher nuclear-spin proposal

The second serious hypothesis is due to Fisher.17 It differs from Orch-OR in the single most important respect: its qubit is a nuclear spin, not an electronic or conformational state. Phosphorus-31 (³¹P, spin-½, ~100 % natural abundance) is, Fisher argues, the only biological element whose nuclear spin can serve as a qubit, transported by the phosphate ion and shielded inside the Posner molecule Ca₉(PO₄)₆.17 Entanglement is created by the chemistry of pyrophosphate hydrolysis (leaving two ³¹P nuclei in a singlet state by quantum-indistinguishability constraints19), and read out via calcium-triggered neurotransmitter release. The proposal makes a concrete, signed, falsifiable prediction: lithium isotopes that differ in nuclear spin (⁶Li vs ⁷Li) should differ in behavioural effect.17 Why a nuclear-spin substrate matters — and where it has been tested — is covered in §4.4.

1.3The origin framing and what it is not

This study began from the double-slit experiment24 and the intuition that “the universe renders an outcome only when observed, to save computation.” In standard physics, “observation” is a physical interaction that correlates a system with its environment — decoherence — not a conscious act.2526 The delayed-choice quantum eraser shows that which-path information, not a mind, governs interference.26 The “renders-on-observation” picture is therefore an analogy to collapse, not a mechanism, and we flag it as such throughout. Its only physics-testable cousin — a discrete-spacetime-lattice signature in cosmic rays — is the subject of probe G4 (§4.5), and is unrelated to consciousness.

2The mathematics of decoherence

The decoherence objection is the foundation of the whole debate, so we set it out in full. The question is purely quantitative: how long can a microtubule quantum state survive in the warm, wet, ion-rich interior of a neuron, and is that long enough for a 25 ms conscious moment? Two published, closed-form decoherence times bracket the dispute. We reproduce each from its own paper’s equations.

2.1Tegmark’s estimate (charge / monopole form)

Tegmark4 modelled the dominant decoherence channel as the nearest environmental ion (e.g. Na⁺) scattering electromagnetically off a charged microtubule “kink” held in spatial superposition. His single-ion result (his Eq. 19) is:

where a is the nearest-ion distance, m and q the environmental-ion mass and charge, T the temperature, N the number of charges in the coherent kink, and s = |r′−r| the superposition separation. In the s ≫ D limit (his Eq. 22) this gives Tegmark’s quoted figure of ~10⁻¹³ s. Critically, Tegmark works in vacuum (ε_r = 1) and takes the separation to be macroscopic on the molecular scale — a conformational flip moves the charge distribution by s ≈ D ≈ 24 nm (“tubulin dimers literally beside themselves”).

2.2The Hagan–Hameroff–Tuszyński rebuttal (dipole form)

Hagan, Hameroff & Tuszyński5 restated Tegmark’s result and re-derived it for a dipole rather than a bare charge (their Eq. 4):

where ε_r is the medium dielectric constant, p the tubulin dipole moment, and Ω an O(1) geometric factor. They obtain 8–9 orders of magnitude more coherence than Tegmark by stacking four changes, which the analysis isolates one at a time:

1. The superposition separation s — the dominant lever. Since τ ∝ 1/s in both formulae, shrinking s by a factor buys the same factor in τ. Hagan argues the London-force dipole flip displaces nuclei by only a Fermi length, and uses s ≈ 0.01–0.1 pm — some 10⁵–10⁹× smaller than Tegmark’s 24 nm. This one choice is most of the dispute.
2. Charge → dipole. Replacing the monopole charge with tubulin’s axial dipole moment p ≈ 337 D (a fifth of the 1714 D total) steepens the ion-distance falloff from a³ to a⁴.
3. Dielectric screening. Tegmark used vacuum (ε_r = 1); Hagan adds conservative cytosol screening, ε_r ≈ 10, for a factor of ~10.
4. Nearest-ion distance a. Hagan places the kink at the tubule surface, a ≈ 14 nm vs Tegmark’s 26 nm; the direction of the effect depends on the a³ vs a⁴ power and is modest.

2.3The order-of-magnitude gap, derived explicitly

Both rival numbers are now placed on one axis against the cognitive threshold:

The shortfall is computed directly. Tegmark’s 25 ms / 3.4 × 10⁻¹⁴ s ≈ 7 × 10¹¹ — about twelve orders of magnitude (from the unrounded 3.4 × 10⁻¹⁴ s anchor). Hagan’s best published figure, 25 ms / 10⁻⁴ s = 250; his typical figure, 25 ms / 10⁻⁵ s = 2500. Because τ ∝ 1/s in both forms, the entire 8–9 orders of advantage that separates the two camps reduces, on inspection, to a single disputed parameter: the superposition separation s. Charge-vs-dipole, dielectric, and ion distance together move τ by only ~10²–10³; the separation moves it by ~10⁵–10⁹.

2.4Can any defensible parameter set reach 25 ms?

We computed the most coherence-favourable physically-defensible corner of the whole parameter space — every knob pushed to its generous-but-justifiable bound, nothing invented:

The result is τ_best ≈ 2.5 × 10⁻² s ≈ 25 ms. At this extreme corner the math touches the 25 ms line (shortfall ≈ 1.0×) — but still falls ~4× short of 100 ms. And the “success” is dominated by the s = 1 Fermi choice: dropping s from 0.1 pm to 1 fm alone buys ~100×; the dielectric 10→80 buys ~8×; the full dipole and the a⁴ term supply the rest. Back s off to a still-tiny 1 pm — still a sub-atomic displacement — and the same maxed-out case gives τ ≈ 2.5 × 10⁻⁵ s, ~1000× short again.

2.5The entanglement-claim numbers and the Fisher coherence argument

Two further sets of numbers are load-bearing elsewhere in the report and are stated here for completeness.

The MRI entanglement-witness magnitudes (§4.3). The classical intermolecular-zero-quantum-coherence (iZQC) signal builds on a dipolar demagnetising time τ_d = 1 / (γ μ₀M₀), with a Curie-law magnetisation M₀. A first-principles computation gives τ_d ≈ 195 ms at 7 T (anchor ~170 ms) and 456 ms at 3 T, scaling as 1/B₀. At Kerskens’s short echo time TE = 5 ms the classical signal is only ~0.36 % of bulk — the same order as Kerskens’s own ~0.13 % estimate, and 40–100× below the observed ~15 %. Warren’s much-quoted 41 % ceiling is relaxation-free and does not apply at 5 ms. Both the classical iZQC signal and a genuine entangled-spin signal carry the same dipolar angular factor |3 cos²θ − 1| and both vanish at the magic angle 54.7° — so the magic-angle null confirms dipolar mediation but does not discriminate classical from quantum.

The Fisher nuclear-spin coherence numbers (§4.4). Nuclear spins are intrinsically ~10⁶–10⁹× better isolated from the environment than electronic states, and second-to-minute room-temperature nuclear coherence is routine in NMR — so Fisher’s starting point is physically grounded in a way Orch-OR’s is not. Fisher estimates t_coh ~ 1 s for a solvated phosphate ion and ~10⁵ s (~1 day) inside a Posner molecule.17 Independent recalculations trend downward: Player & Hore estimate ~37 min from intracluster dipolar coupling;22 Agarwal/Kattnig find the symmetric trimer entanglement decays sub-second but propose a resilient dimer carrier holding entanglement for ~10²–10³ s.21 None of these is an in-vivo measurement. The discriminating prediction rests on the lithium isotopes, which differ by ~14 % in mass as well as in nuclear spin — the confound at the heart of the Fisher verdict (§4.4).

3Methods

This is a computational synthesis and structured literature review, not new wet-lab work. The method is designed to be reproducible by an individual investigator with a workstation and access to the open literature, and is built around a single discipline: separate experimental fact from interpretation, cite every load-bearing claim, and grade evidence conservatively.

3.1The knowledge graph

The literature was ingested into a four-layer pipeline. (1) Ingest: ~35 papers and reviews across quantum consciousness, microtubule biophysics, anaesthesia, quantum biology, and the simulation hypothesis, stored with metadata and abstracts. (2) Extract: a structured record per paper — core claims, mechanism, what was measured, sample/method, result, evidence strength, and which existing claims it supports or contradicts — with a speculation flag where a claim outran its data. (3) Graph: a 47-node, 53-edge knowledge graph of concepts, hypotheses, experiments, evidence, and disagreements, each node carrying an evidence grade (established / contested / speculative / interpretation). (4) Analysis: cross-paper passes to find contradictions, thin evidence, orphan links (concepts that should connect but no paper bridges), and the cheapest test that would settle each — producing a ranked, cited gap list. The explicit honesty guardrail: every output must cite a source; uncited assertions are rejected; retrieved text is treated as data, never as instruction.

3.2The discriminating probes (G1–G6)

The analysis layer placed each probe at a structural pressure-point of a theory and ran the cheapest test capable of separating quantum from classical. Three are computational re-implementations; three are structured literature syntheses:

For the computational probes (G1, G3, G4), formulae were transcribed directly from the source papers’ equations, implemented in Python 3.13 with CODATA SI constants, and gated by reproducing each paper’s own published anchor before any sweep. No parameter was tuned to manufacture a result; where a value is uncertain it is swept, not silently fixed.

3.3Verification

Every probe was independently fact-checked by two adversarial reviewers: one auditing the numbers/formulae (transcription, constants, unit checks, anchor reproduction), the other auditing the conclusion for soundness and honesty (whether the verdict outran the evidence in either direction). All load-bearing numbers were re-checked against the live literature. Verification was not cosmetic: it caught and corrected real errors — for example, in G6 it added a previously-omitted calcium-isotope null result to the evidence-against, corrected the direction of a cited structural paper, and softened a ranking claim the cited review did not support. The corrections are reflected in §4 and in the references.

4Results

4.1G1 — The foundation: the decoherence gap is unbridged

The full derivation is §2. The result is unambiguous: both published decoherence times fall below the 25 ms threshold (Table 1). Tegmark’s figure kills the program outright (~12 orders short); Hagan’s best published figure is still 250× short, and his typical figure 2500× short. The 25-year Tegmark↔Hagan dispute is a genuine 8–9 order-of-magnitude contradiction, but neither side’s number rescues Orch-OR; Hagan’s recomputation makes the theory less dead, not alive. A defensible parameter set can just touch 25 ms (Table 2), but only by adopting the contested Fermi-scale separation in the decoherence formula, and even then it misses 100 ms by ~4×.

G1 verdict Unbridged. The foundational requirement — a quantum state surviving the ~25 ms a thought needs — is not met on the published numbers; the gap closes only on the single most-contested assumption. Scope limit: this covers the closed-form environmental-scattering dispute only; Hagan’s other rebuttal elements (vicinal-water screening, Debye counter-ion layer, dynamical error-correction) are not modelled and remain part of his full feasibility argument.

4.2G2 — The main empirical handle: anaesthesia needs no “quantum”

Anaesthetics reversibly switch consciousness off, so whatever they touch is plausibly part of the substrate — and proponents point at microtubules. A structured effect-size meta-read of 13 anaesthetic–microtubule studies sorted each into microtubule-implicating (classical-compatible) vs quantum-implicating (requiring a quantum observable). The “quantum” inference is not load-bearing.

• The strongest empirical result — Khan 2024,6 in which the microtubule stabiliser epothilone B delays anaesthetic-induced loss of righting reflex (Cohen’s d = 1.9, +~69 s, epoB 0.75 mg/kg, replicated in direction by a 2026 mouse follow-up) — is a clean behavioural result that implicates microtubules and says nothing about any quantum mechanism.
• The two “most quantum-looking” microtubule results dissolve on inspection: Kalra 2023’s anaesthetic effect on exciton diffusion is small (~12–15 %) over ~one tubulin dimer (~6.6 nm), and the authors themselves attribute it to classical altered dielectric screening;7 Craddock’s “quantum channels” compute a classical terahertz dipole oscillation.
• Tally: 0 of 6 microtubule–anaesthesia results require a quantum observable; each is fully explained by a classical microtubule role.

The make-or-break exception is the one genuinely quantum candidate: xenon nuclear-spin isotopes.8 Li et al. 2018 measured loss-of-righting-reflex ED50 in 80 mice and found nonzero-nuclear-spin isotopes (¹²⁹Xe, ¹³¹Xe) ~30 % less potent than spin-0 (¹³²Xe, ¹³⁴Xe) — a larger, cleaner effect than any microtubule result. Honestly graded it still cannot make “quantum” load-bearing, for three verified reasons: it is unreplicated (single 2018 lab, no independent confirmation in eight years); mass-confounded (the nonzero-spin isotopes are also lighter, so a classical “lighter → faster transport → weaker effect” trend reproduces the exact ordering); and non-microtubule (the proposed radical-pair mechanism belongs to the Fisher nuclear-spin camp, not the microtubule program). The parallel lithium effect carries the same mass confound.

G2 verdict Confirmed (high confidence). Anaesthetics clearly act, in part, on microtubules — a classical result. “Quantum” is interpretation, not data. The only genuinely quantum anaesthetic signal (xenon) is unreplicated, mass-confounded, and points away from microtubules.

4.3G3 — The keystone: the MRI signal does not discriminate

The piece most often cited as direct evidence of quantum effects in the conscious brain is Kerskens & Perez 2022:9 an MRI “entanglement-witness” signal — heartbeat-locked, awake-only — read as evidence of macroscopic brain entanglement. The witness theorem is borrowed from quantum gravity: if a mediator entangles two ancillas, it cannot be classical. It is valid — but it only bites if the observed signal is demonstrated to be genuine ancilla–ancilla entanglement. What Kerskens actually has is a multiple-quantum-coherence signal with no single-quantum correlate, and classical intermolecular zero-quantum coherence — the long-range distant dipolar field between water molecules — produces exactly that class of signal with no entanglement at all.10

A toy classical null-model implementing the published iZQC theory at Kerskens’s own sequence parameters (the magnitudes are in §2.5) shows the dispute is narrower than the headlines: at TE = 5 ms the classical iZQC signal is small (~0.36 %, the same order as Kerskens’s own ~0.13 %, far below the observed ~15 %), so Warren’s 41 % ceiling is relaxation-free and does not apply at that echo time. But that only weakens one classical channel; it does not exclude classical iMQC writ large, physiological artifacts, or the missing positive mechanism. The strongest original point, drawn by neither camp: the signal vanishing at the magic angle (54.7°) is cited by Kerskens as proof of dipolar origin, but both the classical and the entangled signal carry the same (3 cos²θ − 1) factor and both vanish there — so it confirms dipolar mediation without discriminating classical from quantum.

G3 verdict Does not discriminate. The keystone “direct evidence” is the signal class that ordinary intermolecular nuclear coherence produces with no entanglement. Status: contested, single-lab, and unreplicated — not debunked, not confirmed. Settling it requires independent multi-centre replication plus either a positive classical forward model reproducing the full signal, or a genuine separability-bound (Bell/PPT-type) inequality the in-vivo signal provably exceeds.

4.4G6 — The Fisher nuclear-spin alternative: best-built, but unconfirmed

Fisher’s proposal (§1.2) is, on the project’s axes, the best-formulated and most genuinely testable quantum-brain hypothesis. Crucially, the nuclear-spin substrate is a real answer to the decoherence problem that sinks Orch-OR (G1): nuclear spins are intrinsically ~10⁶–10⁹× better isolated than electronic states, and second-scale room-temperature coherence is routine in NMR. Fisher is fighting on terrain where long coherence is the norm, not the miracle, and his model makes a concrete, signed, falsifiable isotope prediction.

But the evidence leans unsupported, on four heads:

• The carrier is contested. The 1-day-coherence claim rests on a symmetric, rotationally well-defined Posner trimer. Agarwal/Kattnig find the trimer lacks a well-defined symmetry axis and its inter-trimer entanglement decays sub-second, proposing a resilient dimer instead.21 (Adversarial verification corrected our draft here: Swift et al.20 is actually net supportive of long coherence in the idealised molecule — the skeptical case rests on the carrier and the discriminating tests, not on Swift.)
• The headline coherence time is unsupported. Every more-realistic recalculation trends downward from Fisher’s ~10⁵ s — Player & Hore ~37 min,22 Agarwal sub-second (trimer) to ~10²–10³ s (dimer) — and none is an in-vivo measurement.
• The supporting isotope evidence is mass-confounded. ⁶Li and ⁷Li differ by ~14 % in mass as well as nuclear spin, so the suggestive behavioural and in-vitro effects (Sechzer 1986;23 a 2025 PNAS calcium-phosphate result in the predicted direction18) are consistent with a classical kinetic isotope effect — the same trap G2 found for xenon. The authors’ own framing is maximally cautious (title: a “possible quantum effect”).
• The one clean discriminating test so far returned a null. A 2020 calcium-isotope experiment — ⁴³Ca (nuclear spin) vs ⁴⁰Ca (spinless) in a sevoflurane anaesthesia paradigm — found no isotope dependence (P > 0.9999), against the nuclear-spin prediction.27

G6 verdict Best-built and falsifiable — but speculative and unconfirmed. The nuclear-spin substrate genuinely dodges the decoherence problem, but the symmetric Posner carrier and its long coherence are not established, the supporting isotope evidence is mass-confounded, and the one clean discriminating isotope test is already null. Genuinely alive and testable — not proven, not debunked — and the one quantum-brain idea most likely to be cleanly settled by bench experiments this decade.

4.5G4 — The simulation framing: no lattice signature, and the honest line between physics and analogy

The project’s origin intuition (§1.3) has exactly one physics-testable cousin: Beane, Davoudi & Savage’s prediction that a discrete cubic spacetime lattice — if our universe ran on one, by analogy to lattice QCD — would imprint two signatures on the highest-energy cosmic rays: an energy cutoff at the inverse lattice spacing (E_max ~ 1/b, with b⁻¹ ~ 10¹¹ GeV = 10²⁰ eV), and a cubic-symmetry angular anisotropy.28 Checked against the latest published Pierre Auger and Telescope Array analyses: the high-energy suppression is conventional (the GZK effect plus likely source-acceleration limits), and a lattice cutoff would be degenerate with GZK anyway, so the cutoff can never be evidence for a lattice. The only significant large-scale anisotropy is a single 6.8σ dipole (amplitude d ≈ 0.065, pointing ~115° from the Galactic Centre — extragalactic);29 quadrupole and higher multipoles — exactly where a cubic-lattice imprint would appear — are not significant.

The honest map: Beane’s cubic lattice is falsifiable (and shows no signature, disfavouring the crude-cubic version, though improved/non-cubic/finer lattices evade the test); Bostrom’s trilemma30 is unfalsifiable philosophy; and the “renders-on-observation” framing makes no physical prediction at all — it is an analogy, not a mechanism. The double-slit experiment that started the project shows collapse on interaction, not on conscious observation.

5The contrast: where quantum biology actually is real

This is the essential calibration. Genuine, measurable functional quantum effects in living systems do exist — they are just narrow, and not about the brain:

• Avian radical-pair magnetoreception.1112 Cryptochrome radical-pair spin chemistry gives migratory birds a magnetic compass — the strongest quantum-biology case, graded established in the corpus (the in-vivo causal chain is still being closed).
• Enzyme tunnelling.14 An established functional quantum effect in catalysis.
• Photosynthesis — the cautionary tale. The field’s original flagship “long-lived coherence” result was substantially walked back: room-temperature electronic coherence lasts only ~60 fs (far below the transfer time), and the long-lived beats are largely vibrational, not functional electronic coherence.13

Two lessons follow, both bearing directly on the brain claims. First, real quantum biology operates at femtosecond–picosecond scales — roughly ten to thirteen orders of magnitude shorter than the millisecond cognition Orch-OR needs, which is exactly the gap G1 quantifies. Second, the discipline’s own flagship was over-interpreted and corrected; any microtubule-coherence claim inherits that cautionary precedent — a real spectroscopic signal is not functional coherence. The credible quantum-biology toolkit (2D electronic spectroscopy, magnetic-field-effect assays, isotope substitution) that corrected photosynthesis and confirmed magnetoreception has not been turned on the brain claims with the same discriminating rigour.

6Discussion and verdict

Laid side by side against the structure of the program, the three Orch-OR probes each fail independently:

The Orch-OR-style quantum-brain program is not supported on current evidence. Its foundation (G1), its main empirical claim (G2), and its keystone evidence (G3) each fail, independently, to clear the bar. The wider corpus reinforces the same direction: the original conformational-qubit biochemistry was argued infeasible;15 only the weak/classical Fröhlich regime is biologically feasible;16 and the gravitational-collapse mechanism behind Orch-OR’s “OR” is now experimentally constrained — an underground experiment rules out the parameter-free Diósi–Penrose model.31

But — and this distinction is load-bearing — “not supported” is not “disproven.” None of the probes refutes quantum effects in the brain. G1 shows the gap is unbridged and identifies the single contested assumption that would bridge it; it does not prove that assumption false, and it does not model Hagan’s full feasibility argument. G2 shows the quantum inference is unnecessary for the anaesthesia data, not that nuclear-spin biology is impossible. G3 shows the keystone evidence fails to discriminate and is unreplicated, not that the signal is proven classical — no one has published the positive classical forward model either. The intellectually honest status is a fringe program resting on a foundation that does not clear the threshold, a main empirical claim where “quantum” does no work, and a keystone that does not discriminate — unsupported and contested, but not closed. Fisher’s nuclear-spin proposal is the healthier of the two on the project’s axes, and the one to watch, precisely because it is the most testable — but it too is unconfirmed, with a clean null already on the board.

7Limitations

This report is presented honestly for what it is, because that honesty is what makes it usable. The following limits are not caveats bolted on at the end; they bound every claim above.

• It is a synthesis and review, not original empirical discovery. No new wet-lab data were generated. The contribution is an honest, cited consolidation of the existing literature plus cheap computational probes — not a new measurement of the brain.
• The decoherence probe is closed-form only. G1 re-implements the two published single-ion closed-form decoherence times. It does not model Hagan’s further rebuttal elements (ordered/vicinal-water screening, the Debye counter-ion layer, dynamical topological error-correction, the thermal-equilibrium objection), which remain part of his full biological-feasibility argument.
• The entanglement null-model is a toy, order-of-magnitude envelope. G3 computes the classical iZQC magnitude at Kerskens’s parameters; it is not a multi-compartment in-vivo forward model and cannot, on its own, adjudicate the in-vivo signal. A full forward model is arguably intractable in principle, which is itself part of the dispute.
• The cosmic-ray figures use published collaboration results, drawn schematically. G4 uses Auger/TA published feature energies, indices, and the dipole amplitude/direction; there is no public per-bin flux table for an individual to re-fit, so the spectrum and anisotropy figures are schematic in geometry but anchored to real numbers.
• Conservative grading may understate live possibilities. A real signal with a contested interpretation is graded contested, never established. This deliberately errs toward skepticism; the xenon and lithium isotope effects in particular are real and would, if replicated with the mass confound removed, become genuine quantum biology.
• The work used an AI research collaborator. The literature extraction, computation, and drafting were AI-assisted. Every load-bearing number was independently re-checked on the live literature, and every probe was fact-checked by two adversarial reviewers; the judgement and the standard are the lab’s. Retrieved text was treated as data, not instruction. No citations, numbers, or results were invented; figures that could not be re-derived are attributed to their source, not asserted.

8Conclusion

On current evidence there is no good reason to think quantum mechanics does functional work in the brain. The leading theory (Orch-OR) is not supported — its foundation, its main empirical handle, and its keystone evidence each fail independently. The most testable rival (Fisher) is the one to watch, but it is unconfirmed and already carries a clean null. Real quantum biology is established, but it is narrow, ultra-fast, and not about consciousness. Throughout, the honest claim is the weaker one: “not supported” is not “disproven.” The deliverable is the emperor’s thin clothes mapped precisely, plus the cheapest experiments that would re-tailor them — not a takedown, and not a breakthrough.

9References

All references are real and were cross-checked against the live literature; load-bearing numbers were verified at the source. Where a citation was corrected during adversarial verification (e.g. Player & Hore’s venue, the Sechzer page range, the direction of Swift et al.), the corrected form is given here.

1. Penrose, R. (1989). The Emperor’s New Mind: Concerning Computers, Minds and the Laws of Physics. Oxford University Press. — the non-computability argument and the gravitational objective-reduction origin of Orch-OR.
2. Hameroff, S. & Penrose, R. (1996). Orchestrated reduction of quantum coherence in brain microtubules: a model for consciousness. Mathematics and Computers in Simulation 40(3–4):453–480.
3. Hameroff, S. & Penrose, R. (2014). Consciousness in the universe: a review of the ‘Orch OR’ theory. Physics of Life Reviews 11(1):39–78. doi:10.1016/j.plrev.2013.08.002 — source of the ~25 ms / 100 ms cognitive thresholds.
4. Tegmark, M. (2000). Importance of quantum decoherence in brain processes. Physical Review E 61(4):4194–4206. arXiv:quant-ph/9907009 — the ~10⁻¹³ s decoherence estimate (Eqs. 9/19/22).
5. Hagan, S., Hameroff, S. R. & Tuszyński, J. A. (2002). Quantum computation in brain microtubules: decoherence and biological feasibility. Physical Review E 65(6):061901. arXiv:quant-ph/0005025 — the dipole re-derivation and the ~10⁻⁵–10⁻⁴ s rebuttal.
6. Khan, S., Huang, Y., Timuçin, D., et al. (2024). Microtubule-stabilizer epothilone B delays anesthetic-induced unconsciousness in rats. eNeuro 11(8):ENEURO.0291-24.2024. — the strongest behavioural microtubule–anaesthesia result (Cohen’s d = 1.9, +~69 s, epoB 0.75 mg/kg).
7. Kalra, A. P., Patel, S. D., et al. (2023). Electronic energy migration in microtubules. ACS Central Science 9(3):352–361. — ~12–15 % effect over ~6.6 nm, attributed by the authors to classical dielectric screening.
8. Li, N., Lu, D., Yang, L., et al. (2018). Nuclear spin attenuates the anesthetic potency of xenon isotopes in mice. Anesthesiology 129(2):271–277. — the xenon nuclear-spin isotope effect (n = 80).
9. Kerskens, C. M. & López Pérez, D. (2022). Experimental indications of non-classical brain functions. Journal of Physics Communications 6(10):105001. doi:10.1088/2399-6528/ac94be — the MRI entanglement-witness signal.
10. Warren, W. S. (2023). Comment on ‘Experimental indications of non-classical brain functions’. Journal of Physics Communications 7(3):038001. doi:10.1088/2399-6528/acc4a8 — the classical intermolecular-coherence (iMQC/iZQC) alternative. See also Kerskens & Pérez, Reply, ibid. 7:038002 (doi:10.1088/2399-6528/acc636).
11. Hore, P. J. & Mouritsen, H. (2016). The radical-pair mechanism of magnetoreception. Annual Review of Biophysics 45:299–344.
12. Xu, J., Jarocha, L. E., Zollitsch, T., et al. (2021). Magnetic sensitivity of cryptochrome 4 from a migratory songbird. Nature 594:535–540.
13. Duan, H.-G., Prokhorenko, V. I., Cogdell, R. J., et al. (2017). Nature does not rely on long-lived electronic quantum coherence for photosynthetic energy transfer. PNAS 114(32):8493–8498. — room-temperature electronic coherence ~60 fs; the long-lived beats are largely vibrational.
14. Cao, J., Cogdell, R. J., Coker, D. F., et al. (2020). Quantum biology revisited. Science Advances 6(14):eaaz4888. — enzyme tunnelling as established quantum biology; the photosynthesis walk-back.
15. McKemmish, L. K., Reimers, J. R., McKenzie, R. H., et al. (2009). Penrose–Hameroff orchestrated objective-reduction proposal for human consciousness is not biologically feasible. Physical Review E 80(2):021912.
16. Reimers, J. R., McKemmish, L. K., McKenzie, R. H., et al. (2009). Weak, strong, and coherent regimes of Fröhlich condensation and their applications to terahertz medicine and quantum consciousness. PNAS 106(11):4219–4224.
17. Fisher, M. P. A. (2015). Quantum cognition: the possibility of processing with nuclear spins in the brain. Annals of Physics 362:593–602. arXiv:1508.05929 — the ³¹P nuclear-spin / Posner-molecule proposal.
18. Straub, J. S., Patel, S. D., … Helgeson, M. E. & Fisher, M. P. A. (2025). Evidence for a possible quantum effect on the formation of lithium-doped amorphous calcium phosphate from solution. PNAS 122:e2423211122. — ⁷Li promotes more calcium-phosphate particles than ⁶Li; authors explicitly cautious (“possible”).
19. Fisher, M. P. A. & Radzihovsky, L. (2018). Quantum indistinguishability in chemical reactions. PNAS 115(20):E4551–E4558. — the entanglement-by-chemistry mechanism.
20. Swift, M. W., Van de Walle, C. G. & Fisher, M. P. A. (2018). Posner molecules: from atomic structure to nuclear spins. Physical Chemistry Chemical Physics 20:12373–12380. arXiv:1711.05899 — the idealised S₆-symmetric Posner supports long coherence (corrected: net supportive).
21. Agarwal, S., Kattnig, D. R., et al. (2023). The biological qubit: calcium phosphate dimers, not trimers. Journal of Physical Chemistry Letters. arXiv:2210.14812 — trimer entanglement decays sub-second; dimers preserve it for ~10²–10³ s.
22. Player, T. C. & Hore, P. J. (2018). Posner qubits: spin dynamics of entangled Ca₉(PO₄)₆ molecules and their role in neural processing. Journal of the Royal Society Interface 15:20180494. arXiv:1807.06339 — ~37 min entanglement-decoherence estimate.
23. Sechzer, J. A., Lieberman, K. W., Alexander, G. J., et al. (1986). Aberrant parenting and delayed offspring development in rats exposed to lithium. Biological Psychiatry 21(13):1258–1266. — the seed ⁶Li/⁷Li opposite-direction behavioural datum.
24. Tonomura, A., Endo, J., Matsuda, T., et al. (1989). Demonstration of single-electron buildup of an interference pattern. American Journal of Physics 57(2):117–120. — the double-slit experiment that started the project.
25. Schlosshauer, M. (2005). Decoherence, the measurement problem, and interpretations of quantum mechanics. Reviews of Modern Physics 76:1267–1305. — “observation” is environmental interaction, not a conscious act.
26. Kim, Y.-H., Yu, R., Kulik, S. P., Shih, Y. & Scully, M. O. (2000). Delayed ‘choice’ quantum eraser. Physical Review Letters 84(1):1–5. — which-path information, not a mind, governs interference.
27. Chen, R., Li, N., Qian, H., Zhao, R.-H. & Zhang, S.-H. (2020). Experimental evidence refuting the assumption of phosphorus-31 nuclear-spin entanglement-mediated consciousness. Journal of Integrative Neuroscience 19(4):595–600. doi:10.31083/j.jin.2020.04.250 — the clean discriminating isotope null: ⁴³Ca (nuclear spin) vs ⁴⁰Ca (spinless) gave identical sevoflurane loss-of-righting-reflex ED50 (P > 0.9999), against the ³¹P / Posner nuclear-spin prediction.
28. Beane, S. R., Davoudi, Z. & Savage, M. J. (2014). Constraints on the universe as a numerical simulation. European Physical Journal A 50:148. arXiv:1210.1847 — the cubic-lattice cosmic-ray cutoff + anisotropy prediction.
29. Pierre Auger Collaboration (2025). Large-scale anisotropies of ultra-high-energy cosmic rays measured at the Pierre Auger Observatory. arXiv:2507.19243 — the 6.8σ dipole (d ≈ 0.065, extragalactic); quadrupole and higher not significant. Spectrum: Phys. Rev. D 102:062005 (2020).
30. Bostrom, N. (2003). Are you living in a computer simulation? The Philosophical Quarterly 53(211):243–255. — the trilemma; no falsifiable prediction on its own.
31. Donadi, S., Piscicchia, K., Curceanu, C., Diósi, L., Laubenstein, M. & Bassi, A. (2021). Underground test of gravity-related wave-function collapse. Nature Physics 17:74–78. — rules out the parameter-free Diósi–Penrose gravitational-collapse model.

10Acknowledgments

Computational analysis and drafting assisted by Claude (Anthropic).

---

Provenance & honesty note. Built on a knowledge graph (47 nodes / 53 edges / 35 papers), a ranked gap analysis, and six discriminating probes (G1–G6), each independently fact-checked by two adversarial reviewers (all checks pass). All cited papers are real and correctly cited; all load-bearing numbers were verified on the live literature (the G1 decoherence anchors; Khan d = 1.9; Kalra ~12–15 % / ~6.6 nm; Li 2018 ED50s; Hagan ~10⁻⁴ s vs Tegmark ~10⁻¹³ s; Duan ~60 fs; Kerskens TE = 5 ms / ~15 % vs the iZQC null ~0.36 %; the ⁴³Ca null P > 0.9999). Evidence grades are deliberately conservative; a real signal with a contested interpretation is graded contested, never established. The quantum-brain program is “not supported / contested,” not “debunked”; solid quantum biology (magnetoreception, enzyme tunnelling) is the honest contrast and is graded established. ZA / SAST. The work was done with an AI research collaborator; the judgement and the standard are the lab’s.