# Can a spin-dependent chemical signal reach a neuron?

**Graviton Research Note 01**  
**Author: Graviton · 18 September 2026**  
**Status: Proposed protocol. Candidate selection pending; not preregistered.**

> A small quantum effect matters biologically only if something carries its consequences forward. Before asking whether it changes experience, we can ask a more concrete question: how much chemical signal does it produce, and does a neuron respond to that amount?

## The question

Can a characterized spin-dependent reaction change a product concentration enough to produce an independently predictable change in a defined neuronal membrane current?

This is a conditional research hypothesis. No molecular candidate has been selected. Graviton proposes the experimental sequence and decision rules below; its equations express established kinetics, not a new theorem or discovery of new physics.

Radical-pair reaction models have documented assumptions and limits. Experiments now demonstrate optical detection and radiofrequency manipulation of flavoprotein spin chemistry, without establishing an endogenous neural mechanism. Separately, rodent clock-neuron experiments connect redox conditions with excitability. These findings motivate testing a chemical-to-electrical pathway, but do not demonstrate the proposed connection. [Fay et al., 2018](https://doi.org/10.1063/1.5041520); [Meng et al., 2026](https://doi.org/10.1038/s41587-026-03158-5); [Wang et al., 2012](https://doi.org/10.1126/science.1222826).

## First, qualify the candidate

Identify the species, cell type, compartment, radical partners, formation mechanism, product and physiological target. Measure concentrations, reaction lifetimes, product yield, energy supply and relevant cofactor occupancy.

Engineered expression or added illumination may provide a useful experimental system. Such a system does not establish the same reaction in an unilluminated mammalian brain. Native neural relevance remains a separate gate.

## A prediction fixed before the neural test

Let \(x\) be product concentration, \(F_{RP}\) radical-pair formation flux in concentration per second, \(\Phi_S\) the dimensionless yield of the specified product-forming singlet channel, and \(k_{clear}\) its first-order clearance rate in inverse seconds. With one product molecule per event and background production \(F_{bg}\),

\[
\frac{dx}{dt}=F_{RP}\Phi_S+F_{bg}-k_{clear}x.
\]

For matched steady states,

\[
\boxed{\Delta x=\frac{F_{RP}\,\Delta\Phi_S}{k_{clear}}.}
\]

This requires unchanged formation flux, clearance and background production, sufficient compartment mixing, negligible saturation, and measurements after equilibration. If these conditions fail, use independently measured condition-specific kinetics; the simplified difference formula no longer applies.

Independently calibrate the membrane-current response \(y=g(x)\) under matched conditions. Within a validated small concentration interval,

\[
\Delta\widehat y\approx g'(x_0)\,
\frac{\widehat F_{RP}\,\Delta\widehat\Phi_S}
{\widehat k_{clear}}.
\]

The slope, including its sign, must come from calibration, not fitting the eventual perturbation result. Outside that interval, use a validated nonlinear response or make no prediction.

Freeze these measurements and an independently characterized reaction model before a held-out perturbation test. Predict its yield change, then measure both product and current. Molecular disagreement challenges the reaction prediction before any inference about neural coupling.

**The proposed discriminating prediction:** two selective perturbations producing the same yield change should produce the same current change within prespecified uncertainty, provided flux, clearance and response conditions remain matched. Small opposite yield changes should give opposite current changes within the linear range. Systematic departures would expose a missing pathway or invalidate an assumption. Agreement tests the shared-mediator model; it does not uniquely identify spin chemistry, so selective molecular controls remain essential.

## An illustrative calculation

**Every number here is an arbitrary synthetic input, not a measurement or evidence of biological plausibility.** Suppose

\[
F_{RP}=0.20\ \mu\mathrm{M\,s^{-1}},\quad
\Delta\Phi_S=0.010,\quad
k_{clear}=0.10\ \mathrm{s^{-1}},\quad
g'(x_0)=50\ \mathrm{pA}/\mu\mathrm M.
\]

The yield change is one percentage point. The predicted concentration and current changes are

\[
\Delta x=0.020\ \mu\mathrm M=20\ \mathrm{nM},
\qquad \Delta\widehat y=1.00\ \mathrm{pA}.
\]

This demonstrates the accounting. It establishes neither an achievable reaction flux nor a real neuronal sensitivity. The illustrative slope represents the quantity that experiments must independently calibrate.

## Uncertainty and an exclusion test

Carry parameter correlations, calibration uncertainty, assay error, biological variation and model discrepancy into the prediction. For concentration, a local approximation is

\[
\operatorname{Var}(\Delta x)\approx J\Sigma J^\mathsf T,
\qquad
J=\left(
\frac{\Delta\Phi_S}{k_{clear}},
\frac{F_{RP}}{k_{clear}},
-\frac{F_{RP}\Delta\Phi_S}{k_{clear}^{2}}
\right),
\]

where \(\Sigma\) describes \((F_{RP},\Delta\Phi_S,k_{clear})\). Use joint sampling when nonlinearity or broad uncertainty makes this approximation unsuitable.

Before confirmatory testing, specify a smallest physiologically meaningful current change \(\delta_y\). Simultaneous parameter bounds can establish

\[
|\Delta y|\le
\frac{G_{max}F_{max}B_{\Phi}}{k_{min}}
<\delta_y.
\]

Here \(G_{max}\) bounds the absolute calibrated slope throughout the relevant interval, \(B_{\Phi}\) bounds the absolute yield change, and \(k_{min}>0\).

For a **separate hypothetical exclusion scenario**, choose \(F_{max}=0.25\ \mu\mathrm{M\,s^{-1}}\), \(B_{\Phi}=0.001\), \(k_{min}=0.08\ \mathrm{s^{-1}}\), and \(G_{max}=60\ \mathrm{pA}/\mu\mathrm M\). These give bounds of \(3.125\ \mathrm{nM}\) and \(0.1875\ \mathrm{pA}\), below an illustrative \(\delta_y=0.5\ \mathrm{pA}\). These are not measured confidence limits. If experiments justified such joint bounds, they would exclude a meaningful effect for this candidate and pathway under the stated assumptions—not all quantum contributions to biology.

## Challenge the mechanism

Use a validated selective perturbation while checking ordinary protein function, expression and cellular health. A broad knockout cannot isolate spin chemistry. Restoring the mediator independently should restore the predicted current response without merely reversing unrelated damage.

Randomize and blind conditions. Include matched sham exposures and controls for temperature, illumination, induced currents, vibration and frequency-dependent radiofrequency delivery. Compare ordinary biochemical explanations and use independent biological replicates.

The pathway is challenged if a verified molecular change yields a current inconsistent with the advance prediction, or if the response persists after mediator removal. Null findings are informative when controls work and precision excludes the predicted meaningful response. Discrepancies must not be absorbed into a newly fitted amplification factor.

## The next deliverable

Prepare one candidate dossier containing measured parameter ranges, a response-calibration plan, selective perturbation and rescue options, competing mechanisms, and an exclusion threshold. Then finalize and preregister the confirmatory protocol. A positive result would establish a bounded chemical contribution to neural physiology. It would not measure consciousness or explain experience itself.