Hearing the light: stray-field noise from the emergent photon in quantum spin ice

The paper is largely internally coherent within its stated effective-field-theory framework. The central logical chain—Maxwell action for the emergent photon (Eq. 2), derivation/motivation of two possible boundary conditions (Eqs. 3–4), cavity/waveguide mode quantization (Eq. 6 and Supplementary Eqs. 21–29), and computation of resulting magnetic stray-field noise (Eq. 7 and Supplementary Eqs. 30–65)—is consistent. In particular, the sharp distinction between the two boundary classes is used consistently throughout: insulating boundaries imply no external stray magnetic noise, while superconducting boundaries permit it. The claim of exactly zero stray field for insulating boundaries is supported both physically and mathematically by Supplementary Eq. (39), and this conclusion is applied consistently in the cavity and thin-film sections. There are, however, a few internal tensions. First, the main text states that from e_parallel = 0 and the homogeneous Maxwell equation one obtains b_perp = 0; this is not an instantaneous pointwise implication without additional assumptions, but only emerges in the long-wavelength/local-boundary limit made explicit later in the supplement via penetration-depth arguments. So the main-text derivation is somewhat compressed relative to the more careful supplement. Second, the discussion of the “natural” boundary conditions from assumptions (i)–(iii) is not a full classification theorem: ruling out e_parallel ∝ b_parallel by time-reversal and derivative terms still leaves open more general nonlocal/frequency-dependent boundary relations unless one assumes strict locality and homogeneity. The authors acknowledge modified spectra when assumptions fail, which helps preserve consistency. Third, the parameter conversion for g and the dimensional statements around Eq. (7) are used consistently numerically, but the precise normalization conventions linking emergent electric field to physical magnetization are not fully derived in the text, so consistency here relies on cited prior work. Overall the manuscript is logically coherent, but some key boundary-condition arguments are more heuristic in the main text than fully closed.

Several central calculations are mathematically sound in form: the cavity-mode expansion and normalization in the SM (Eqs. (21)–(29)) are consistent with a free-field Hamiltonian (Eq. (29)); the Fourier-space proof of vanishing stray field for insulating BC is a valid transversality argument: the integrand in SM Eq. (39) is proportional to (η·k)=0, yielding B=0. The Ewald splitting (SM Eqs. (40)–(50)) is standard and correctly organizes the short/long range pieces, at least at the level of identities and kernel transforms. However, there are notable mathematical issues/omissions that prevent a higher score: 1) Dipole kernel Fourier transform: SM Eq. (35) states \tilde H_{μν}(q)=μ0 q_μ q_ν/q^2. This is the Fourier transform of the distribution −μ0 ∂_μ∂_ν(1/4πR) but it typically requires adding/subtracting a contact term proportional to δ_{μν} (and specifying conventions for excluding R=0 / principal value). Since the probe point is outside the sample and the integrals avoid R=0, the contact term may not contribute, but the paper does not state the needed distributional caveat. The cancellation in Eq. (39) likely survives, but the transform as written is incomplete without conditions. 2) Units/dimensions are not consistently tracked. In Eq. (2) the prefactor ℏ/(8πα′ v) implies ε′=ℏ/(4πα′ v) (as used in SM), but later μ′ is defined inconsistently: SM below Eq. (12) gives μ′=4πα/(ℏ v) (uses α not α′). This looks like a typo, but it affects the stated boundary relation n×b=−μ′ j (Eq. (12)) and the derived penetration depths (Eq. (16)–(19)). If α is truly meant, then earlier identifications of ε′ and μ′ as inverses are inconsistent; if it is a typo and α′ is intended, then fixable. 3) Boundary-condition derivation from energy conservation/time reversal: in the main text, the step “(e×b)_⊥=0 implies either e∥∝b∥ or |e∥||b∥|=0” is not a logically exhaustive classification. In general, vanishing normal Poynting component only requires (e∥×b∥)·n=0, i.e. e∥ and b∥ parallel (including one being zero), but the subsequent exclusion of e∥=u b∥ by time reversal and exclusion of derivative BC by assumption (iii) is plausible yet not presented as a rigorous theorem. This is more a gap in derivation than an error, but it weakens mathematical rigor. 4) In the thin-film section, the kernel H_{μν,k}(z,ω) (SM Eq. (57)) uses κ=\sqrt(|k|^2−ω^2/c^2) with c the true speed of light; earlier the analysis assumes quasi-static ω d/c≪1. If quasi-static is intended, κ≈|k| should be used consistently or the regime of validity of the retarded kernel should be specified. As written, it mixes magnetostatic and retarded forms without a clear approximation statement. 5) Some normalization factors appear inconsistent between main text Eq. (7) and SM Eq. (33): main text has \sqrt(4πα′ v ω_s/V) whereas SM uses \sqrt(ℏ ω_k/(ε′ V)) with ε′=ℏ/(4πα′ v), which are algebraically consistent, but the main text uses Bμs( r )=μ0 g \sqrt(4πα′ v ω_s/V) ∫H W whereas SM has μ0 g \sqrt(ℏ ω/(ε′ V)) ∫H W; since ℏ/ε′=4πα′ v, these match. This is valid but only if the same W normalization (order-1, dimensionless) is used; the paper assumes it without explicitly confirming in main text. In summary: the core mode quantization and the insulating-boundary cancellation are mathematically plausible and largely correct, but there are definitional/transform caveats and at least one apparent coupling typo that affect boundary equations and penetration depths.

The paper makes clear, experimentally testable predictions. Its central claim is that stray-field magnetometry can directly detect emergent photon modes in quantum spin ice through discrete spectral features and spatially structured magnetic noise. The predictions are not merely qualitative: the authors specify expected frequency scales for finite cavities (~100 kHz for ~100 μm samples), thin-film thresholds at ω = nπv/Lz, probe-distance dependence of spectral sharpness, and order-of-magnitude signal levels in experimentally accessible units. A particularly strong falsifiability feature is the sharp distinction between two boundary-condition classes: for ideal 'insulating' boundaries, the stray-field noise from emergent photons is exactly zero outside the sample, whereas for 'superconducting' boundaries it is finite and structured. This gives a strong differential experimental test. The work is also falsifiable because the proposed signatures differ from generic broadband magnetic noise expectations: one should observe cavity-mode quantization, mode-dependent spatial patterns, and geometry-controlled spectra. Failure to observe these features in candidate QSI systems under the stated temperature/frequency/sample-size conditions would weigh against the specific mechanism. The paper does not give a dedicated section titled 'falsification criteria,' but the criteria are implicit and concrete. One limitation is that uncertain microscopic boundary conditions leave room for null or weakened signals to be attributed to non-ideal boundaries rather than absence of the Coulomb phase, which softens decisiveness in practice. Still, the predictions themselves are highly testable and quantitatively stated.

The paper is generally clear, well organized, and readable for a graduate-level condensed matter or AMO audience. The narrative structure is effective: it starts from the experimental problem, motivates the use of stray-field noise, introduces the emergent Maxwell description, develops boundary conditions, and then presents predictions for finite cavities and thin films before discussing microscopic boundary origins and feasibility. The notation is mostly consistent, especially the explicit distinction between true electromagnetic fields (capital letters) and emergent fields (lowercase), which helps prevent confusion. The strongest clarity points are the physical interpretation of the two boundary conditions and the link between those conditions and observable noise. The figures, as described, appear well chosen to communicate both spectral and spatial predictions. That said, some sections are dense and move quickly, especially the derivation and physical justification of the boundary conditions and the transition from microscopic pyrochlore boundary terminations to coarse-grained effective boundary phases. Readers outside the immediate subfield may also find the use of 'superconducting' and 'insulating' potentially confusing despite the stated analogy, since these refer to emergent boundary behavior rather than ordinary electronic phases. In addition, the dependence of conclusions on idealized boundary assumptions could be signposted more explicitly in the main text. Overall, the communication is strong, but a few conceptual dependencies could be made more explicit.

This work demonstrates significant novelty in both approach and synthesis. While quantum spin ice and emergent photons are known concepts, the authors propose a genuinely new experimental approach using stray-field magnetometry to detect these photons. The key novel insights include: (1) the realization that boundary conditions fundamentally determine whether stray fields exist, (2) the prediction of cavity quantization effects in finite QSI samples creating discrete spectral features, and (3) the connection between emergent electrodynamics and measurable magnetic noise patterns. The framework synthesizes conventional physics (Maxwell equations, dipole fields) with emergent phenomena to generate novel predictions about the spatial and spectral structure of magnetic noise. This represents a new way to probe quantum spin liquids experimentally.

The paper is largely complete for its stated goal: proposing and theoretically characterizing stray-field magnetometry as a probe of emergent photons in quantum spin ice. The central quantities are defined, the physical setup is clear, and the main claims are supported by explicit modeling for both finite cavity and thin-film geometries. The paper states its assumptions about boundary physics, derives two natural boundary-condition classes, shows how these affect the mode structure, and connects the resulting spectra to experimentally measurable NV-center T1/T2 observables. Important edge behavior is addressed, most notably the striking limiting case that insulating boundary conditions produce exactly zero external stray-field noise, while superconducting boundaries yield observable signals. The supplement fills in several derivational gaps from the main text, including boundary scalar-field modeling, cavity quantization details, proof of vanishing noise for insulating boundaries, Ewald summation methodology, and thin-film formulas. That said, the work is not maximally complete. Several key steps are deferred to the Supplementary Material rather than sketched in the main text, especially the derivation of the boundary conditions and the practical details of the mode-counting and field convolution. More importantly, the treatment of realistic boundaries is qualitative: the authors acknowledge mixed or nonideal boundary conditions, mesoscopic screening lengths, and microscopic termination dependence, but do not provide a parametrized interpolation showing how robust the predicted signatures are away from the ideal limits. The claim that generic deviations would reduce signal only by an order-one geometric factor is plausible within the paper's framework but not demonstrated in detail here. Thus the paper is strong and mostly self-contained, but some practical edge cases and robustness analyses remain only partially developed.