{"id":"69f9404a-b8fe-41a8-984e-7a40f1835f77","arxiv_id":"1909.01912","paper_version":3,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":4,"one_line_summary":"A localized single-photon excitation in a 2D atomic array can be converted, via a subradiant collective mode, into a highly collimated photon beam.","lead":"This paper simulates a protocol where a single photon, initially parked on a few atoms in a two-dimensional cold-atom array, spreads into a long-lived collective subradiant mode covering the whole array. A controlled Zeeman shift then releases it as a photon beam aimed perpendicular to the array, which could serve as a directional link between quantum computing modules.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the central transfer-and-release claim is supported by the numerics; the AFM release is argued by symmetry, and the polarization selection rules make that argument sound.","rationale":"The paper's central claim is a theoretical proposal. The single-excitation master equation is standard, and the numerics are transparent. The uniform-mode route is fully simulated, including the release and far-field collimation. The AFM route's initialization requires alternating Zeeman shifts; this is a demanding experimental control but not a logical inconsistency, and the paper offers plausible implementation paths. The only unsimulated step is the AFM release, but the Zeeman coupling connects out-of-plane x-polarization to in-plane y-polarization only; the alternating Δ provides the missing momentum to connect the corner-mode AFM state to the zero-momentum PI mode, while the uniform PP mode is not coupled because it is x-polarized. Thus the two-mode model's omission is justified. The low success probability is explicit and does not undercut the claim. I therefore agree with the reader's ACCEPT and would not change the verdict; running the AFM release simulation would be a worthwhile completeness check.","tokens_in":17212,"tokens_out":25238,"duration_ms":243324,"concrete_test":"Run a full 31x31 lattice simulation of the AFM storage protocol (d=0.45λ, initial localized checkerboard excitation) and then apply the alternating Zeeman pulse Δ=3.3γ for 0.5/γ; extract the far-field pattern. If it shows the same narrow θ_max as the uniform release and a comparable fraction of the remaining excitation is emitted, the AFM-release claim is fully confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"After good-faith review, I find no load-bearing flaw in the central claim. The single-excitation master equation, the biorthogonal mode decomposition, and the full-lattice release simulations support the claim that a localized excitation can be redeployed into a delocalized subradiant mode and then released as collimated light. The reader's flagged concern about site-dependent alternating Zeeman shifts affects experimental realizability of the AFM route, not theoretical validity; the paper proposes concrete implementation paths (standing-wave AC Stark field, hyperfine-state-dependent lattices). The one step that is not directly simulated is the AFM release, stated in Sec. II E as 'could also be released in a similar manner.' I checked the symmetry: the Zeeman term couples x and y polarizations site-locally, so an alternating pattern couples the AFM out-of-plane (x, corner momentum) mode only to the uniform in-plane (y, zero momentum) mode; the uniform out-of-plane mode is x-polarized and is not directly coupled. Hence the two-mode model is justified. The low single-photon success probability (~3%) is explicitly disclosed and does not contradict the 'probabilistic' claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes and analyzes a protocol in a two-dimensional square array of cold atoms with a J=0 to J'=1 transition. Starting from a single-photon excitation localized on a few central atoms, the authors show numerically that the excitation can evolve into a delocalized, strongly subradiant collective eigenmode: either the uniform out-of-plane mode or the antiferromagnetic out-of-plane mode. The stored excitation is then released by turning on a Zeeman level shift that couples the subradiant mode to the uniform in-plane bright mode, producing photon emission collimated along the axis perpendicular to the array plane. The central results are obtained by exact numerical solution of the single-excitation dipole-dipole master equation, complemented by an eigenmode analysis, a two-mode model for the release dynamics, and a disorder study based on stochastic position fluctuations.","tokens_in":17446,"tokens_out":3295,"duration_ms":38856,"significance":"If the results hold, this is a valuable and concrete protocol for converting an omnidirectionally emitting localized excitation into directional, collective emission, with potential application to modular quantum information architectures. The paper's strengths include the use of exact numerical integration of the many-atom master equation, clearly stated parameters with no fitting to the target output, a direct demonstration of subradiance-protected spreading for two distinct modes, quantitative far-field collimation data for several lattice sizes, and an explicit treatment of position fluctuations that shows robustness of the storage enhancement. The authors also honestly disclose the modest single-photon success probability (~3%) of the transfer, which is consistent with the probabilistic nature of the scheme.","major_comments":[],"minor_comments":[{"comment":"The release of the antiferromagnetic mode is stated rather than demonstrated: the text says the stored AFM mode 'could also be released in a similar manner' and numerical results are shown only for the uniform mode. Since the two-mode model and the staggered Zeeman coupling make the claim plausible, this is not a blocking issue, but a concise symmetry argument or a sample release simulation would remove the only extrapolation in the protocol.","section":"II E"},{"comment":"The occupation measure L_j is defined using v_j^T b(t) even though H' is non-Hermitian and its eigenvectors are not mutually orthogonal in the usual sense. Please clarify whether a biorthogonal convention is intended and how the normalization in the denominator is fixed; this would help readers reproduce the occupation plots.","section":"II A, Eq. (15)"},{"comment":"The caption contains a typo: 'Here the the 31×31 lattice' should read 'Here the 31×31 lattice'.","section":"Fig. 3 caption"},{"comment":"The statement that the delocalized uniform mode 'quickly grows to 100% of the remaining excitation' could be misread as unit occupation; consider adding a brief sentence clarifying that this refers to the normalized occupation measure among the surviving excitation, not the total initial probability.","section":"II D 2"}],"recommendation":"minor_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper is one of the cleaner theory proposals for using subradiant collective modes in atomic arrays as a quantum interface. The central claim is supported by exact single-excitation master equation numerics with the physical dipole-dipole kernel, no fitted parameters, and the authors are upfront about the modest success probability. I agree with the reader's ACCEPT verdict.\n\nThe new bit is the 2D localized-excitation protocol. Earlier work from this group and others looked at driven arrays or 1D chains; here they show that a localized excitation at the center of a 2D lattice evolves into one of two subradiant eigenmodes — the uniform out-of-plane mode or the antiferromagnetic mode with alternating phases — and that turning on a Zeeman splitting then couples the dark mode to a bright in-plane mode that radiates as a collimated beam normal to the array. The far-field calculations make the collimation concrete, and the disorder study (position fluctuations) shows the protection survives realistic lattice jitter.\n\nThe soft spots are real but minor. The antiferromagnetic route requires site-dependent alternating Zeeman shifts, which is a significant experimental ask. The paper proposes standing-wave AC Stark fields or hyperfine-state-dependent lattices, but no demonstration. The uniform-mode route is more robust. Also, the AFM release is not explicitly simulated; the claim is argued by symmetry and the two-mode model. I checked that argument and it is sound: the alternating Zeeman pattern couples the AFM out-of-plane mode only to the uniform in-plane mode, so the two-mode description is justified. The low success probability (~3% for a nine-atom initial excitation) is a real limitation of the protocol but it is disclosed and the authors show it improves with larger initial excitations.\n\nThe citation pattern is appropriate: the prior work on storing light in subradiant modes and on 1D subradiance-enhanced transport is fully cited, and the present contribution is distinct. There are no invented entities or fitted outputs; the target modes are eigenmodes of the same Hamiltonian, but their dominance at long times follows from their subradiance, not from an imposed constraint.\n\nI would bring this to the reading group and I would cite it. It is a careful, honest theoretical contribution that deserves a serious referee and likely publication after minor revisions. The main thing to check in review — the experimental feasibility of the AFM initialization — is already flagged by the authors.\n\nBest,\n[Your name]","headline":"A credible numerical demonstration that a localized single photon in a 2D atomic array can be stored in a subradiant collective mode and released as a collimated beam; the AFM route has an experimental caveat but the physics is sound.","tokens_in":17969,"tokens_out":2433,"would_cite":true,"duration_ms":23588,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper shows that a localized single-photon excitation in a 2D atomic array can be transferred into a delocalized subradiant collective mode and then released as a photon beam collimated along the axis normal to the array plane.","keywords":["subradiance","atomic arrays","collective eigenmodes","single-photon storage","directional emission","dipole-dipole interactions","antiferromagnetic collective mode","Zeeman level shifts"],"falsifier":"Prepare a $31\\times31$ square array of atoms with spacing $d=0.75\\lambda$, excite the central nine atoms to an in-phase out-of-plane polarization, wait $\\gamma t\\approx800$, then apply Zeeman splitting $\\Delta=3.3\\gamma$ for $0.5/\\gamma$ and measure the angular distribution of the emitted light: the central claim fails if the emission is not sharply peaked near the array normal with $\\theta_{\\mathrm{max}}\\approx0.05(\\pi/2)$, rather than omnidirectional.","tokens_in":17014,"feed_emoji":"⚛️","tokens_out":11912,"duration_ms":105409,"temperature":0.7,"pith_summary":"This paper establishes a protocol in which a single photon initially localized on a few atoms of a two-dimensional square array---and therefore free to radiate in all directions---spreads into a delocalized collective excitation whose decay is suppressed by subradiance. The authors identify two target eigenmodes that can store the excitation: a uniform out-of-plane mode with all dipoles in phase, and an antiferromagnetic mode with neighboring dipoles $\\pi$ out of phase, the latter protected by lying outside the light cone at spacings below $\\lambda/\\sqrt{2}$. Turning on a Zeeman level shift couples the stored dark mode to a bright in-plane mode, so the photon is emitted only within a narrow cone around the axis perpendicular to the array. The result is a free-space, cavity-free conversion of local omnidirectional emission into effectively one-dimensional photon propagation, which could serve as a coherent link between stages of a quantum information architecture.","feed_headline":"An atomic array turns one photon into a collimated beam","feed_subtitle":"A localized excitation is stored in a long-lived collective mode, then released only along the plane's normal axis.","key_machinery":"The machinery is the collective-mode description of the non-Hermitian coupling matrix $H'$ whose entries come from the dipole radiation kernel $G'$ between every pair of atoms; its eigenvectors are the collective excitation modes, each with a resonance shift and a linewidth. Subradiant eigenmodes have very small linewidths and therefore dominate after the transient modes decay. Zeeman level shifts break the isotropy of the $J=0\\to J'=1$ transition and rotate out-of-plane polarization into in-plane polarization; this is how the localized excitation is imprinted with the correct phase pattern and how the stored excitation is later coupled to the bright in-plane mode. The release step is summarized by a two-mode model, Eqs. (24), for the dark out-of-plane and bright in-plane modes.","core_discovery":"On the paper's own terms, the central discovery is that subradiant collective eigenmodes of a two-dimensional atomic lattice can act as protected transport channels: an excitation placed on roughly 1% of the atoms decays into a mode that extends over the entire array, and because that mode couples weakly to radiation, it survives long after all other components have decayed. The uniform out-of-plane mode has linewidth scaling roughly as $N^{-0.9}$, while the antiferromagnetic mode below the light line is far more protected, with linewidth scaling roughly as $N^{-3}$; a localized excitation with a checkerboard phase pattern feeds this mode. Releasing the stored photon by Zeeman splitting transfers the population to the uniform in-plane bright mode, and the resulting far field is a zeroth-order Bragg peak: for a $31\\times31$ array the 99%-containment angle is $\\theta_{\\mathrm{max}}\\approx0.05(\\pi/2)$, dropping to $\\approx0.02(\\pi/2)$ for $71\\times71$, tending to pure one-dimensional propagation in the infinite-array limit.","pith_inferences":["Because the same array can be read out with either a uniform or an alternating Zeeman pattern, the protocol could act as a controllable router: the choice of shift pattern selects which protected mode is coupled to the bright mode, releasing photons with different spatial phase profiles from the same stored state.","Since the single-excitation dynamics is formally equivalent to classical coupled dipoles, the collimation itself is a classical wave effect; a quantum advantage would require multi-photon or entangled states, which the paper leaves as future work.","The two-mode release step is structurally an EIT analogue in which the coupling laser is replaced by a global level shift, suggesting the scheme could be extended to stop or delay weak coherent pulses rather than single photons, with the array as a reconfigurable quantum memory.","A direct experimental test could use a clock transition with small $d/\\lambda$, such as the $^3P_0\\to{}^3D_1$ transition of $^{88}$Sr mentioned in the paper, where $d/\\lambda\\approx0.08$ puts the AFM mode deep in the subradiant regime."],"forward_implications":["For a $31\\times31$ array at $d=0.75\\lambda$, an in-phase out-of-plane excitation on the central nine atoms evolves so that the uniform out-of-plane mode carries essentially all of the surviving excitation after times of order $800/\\gamma$, despite having only about 3.5% of the initial occupation.","Releasing from the uniform mode with $\\Delta=3.3\\gamma$ for $0.5/\\gamma$ produces emission collimated to $\\theta_{\\mathrm{max}}\\approx0.05(\\pi/2)$ for $31\\times31$ and $\\approx0.02(\\pi/2)$ for $71\\times71$; in the infinite-array limit the radiation is purely along the array normal.","The antiferromagnetic mode at $d=0.45\\lambda$ lies below the light line and has a linewidth much smaller than the uniform modes, so it offers the longest storage; the cost is an initial checkerboard phase pattern over the excited atoms.","Because the initial excitation can be a single photon, the occupation measure directly gives the probability that the photon reaches the target mode: about 1 in 30 realizations for the antiferromagnetic example and somewhat lower for the uniform example.","Finite position fluctuations around lattice sites degrade the lifetime, but the out-of-plane excitation still decays far more slowly than an in-plane localized excitation, so the protection does not require perfectly pinned atoms."],"supporting_citations":[{"why":"Supplies the earlier demonstration that Zeeman splitting couples in-plane and out-of-plane collective modes and defines the occupation measure used to track the target mode.","marker":"[27]"},{"why":"Provides the two-mode model and the analytic large-N linewidth for the uniform in-plane mode that the release step relies on.","marker":"[29]"},{"why":"Establishes the photonic band structure of 2D atomic lattices, which explains why the antiferromagnetic mode at q=(pi/d, pi/d) cannot decay below the light line.","marker":"[22]"},{"why":"Gives the cooperative-transparency framework and the expression for the uniform in-plane mode linewidth used in Eq. (23).","marker":"[24]"},{"why":"Shows exponential improvement in photon storage using subradiance in atomic arrays, used as the benchmark for the lifetime scaling.","marker":"[32]"},{"why":"Provides the theory of subradiant states of a one-dimensional atom chain whose N^{-3} scaling is quoted for the AFM-mode linewidth.","marker":"[48]"},{"why":"Proposes the free-space photonic quantum link architecture that this protocol would connect to as a quantum-information stage.","marker":"[36]"},{"why":"Provides the quantum master equation formalism for many-atom dipole-dipole interactions on which the single-excitation dynamics is built.","marker":"[72]"},{"why":"Supplies the dipole radiation kernel used to compute the interaction matrix and the far-field emission.","marker":"[73]"},{"why":"Shows how Zeeman level shifts can be controlled with microwave or AC-Stark dressing, the control mechanism for initialization and release.","marker":"[76]"}],"fun_headline_variants":["Atomic array funnels a photon into a narrow beam","Subradiant atomic array turns one photon into a beam","Atomic array collimates a single photon via subradiance","Protected atomic modes steer photon emission along one axis","Long-lived atomic mode directs a photon's path"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The protocol assumes that a localized single excitation can be created with a prescribed phase pattern---including, for the longest-lived antiferromagnetic mode, alternating $\\pi$ phases on neighboring sites via site-dependent Zeeman shifts---so if that level of single-site control cannot be achieved with sufficient fidelity, the protected spreading and collimated release cannot occur.","fun_headline_variants_meta":{"raw":{"variants":["Atomic array funnels a photon into a narrow beam","Subradiant atomic array turns one photon into a beam","Atomic array collimates a single photon via subradiance","Protected atomic modes steer photon emission along one axis","Long-lived atomic mode directs a photon's path"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000481,"raw_usage":{"total_tokens":2384,"prompt_tokens":954,"completion_tokens":1430,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":570,"completion_tokens_details":{"reasoning_tokens":1367}},"tokens_in":570,"tokens_out":1430,"duration_ms":11729,"temperature":1.0,"reasoning_tokens":1367,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:04:52.989269+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Prepare a $31\\times31$ square array of atoms with spacing $d=0.75\\lambda$, excite the central nine atoms to an in-phase out-of-plane polarization, wait $\\gamma t\\approx800$, then apply Zeeman splitting $\\Delta=3.3\\gamma$ for $0.5/\\gamma$ and measure the angular distribution of the emitted light: the central claim fails if the emission is not sharply peaked near the array normal with $\\theta_{\\mathrm{max}}\\approx0.05(\\pi/2)$, rather than omnidirectional.","supporting_citations":[{"cited_title":"Metamate- rial transparency induced by cooperative electromagnetic interactions,","cited_arxiv_id":null,"evidence_quote":"Supplies the earlier demonstration that Zeeman splitting couples in-plane and out-of-plane collective modes and defines the occupation measure used to track the target mode."},{"cited_title":"Enhanced Optical Cross Section via Collective Coupling of Atomic Dipoles in a 2D Array,","cited_arxiv_id":null,"evidence_quote":"Provides the two-mode model and the analytic large-N linewidth for the uniform in-plane mode that the release step relies on."},{"cited_title":"Resonant metalenses for breaking the diffraction barrier,","cited_arxiv_id":null,"evidence_quote":"Establishes the photonic band structure of 2D atomic lattices, which explains why the antiferromagnetic mode at q=(pi/d, pi/d) cannot decay below the light line."},{"cited_title":"Controlled manipulation of light by cooperative response of atoms in an optical lattice,","cited_arxiv_id":null,"evidence_quote":"Gives the cooperative-transparency framework and the expression for the uniform in-plane mode linewidth used in Eq. (23)."},{"cited_title":"Interaction of light with planar lattices of atoms: Reflection, transmis- sion, and cooperative magnetometry,","cited_arxiv_id":null,"evidence_quote":"Shows exponential improvement in photon storage using subradiance in atomic arrays, used as the benchmark for the lifetime scaling."},{"cited_title":"Optimiza- tion of photon storage fidelity in ordered atomic arrays,","cited_arxiv_id":null,"evidence_quote":"Provides the theory of subradiant states of a one-dimensional atom chain whose N^{-3} scaling is quoted for the AFM-mode linewidth."},{"cited_title":"Phase-imprinted multiphoton subradiant states,","cited_arxiv_id":null,"evidence_quote":"Proposes the free-space photonic quantum link architecture that this protocol would connect to as a quantum-information stage."},{"cited_title":"Subradiance-enhanced excitation transfer be- tween dipole-coupled nanorings of quantum emitters,","cited_arxiv_id":null,"evidence_quote":"Provides the quantum master equation formalism for many-atom dipole-dipole interactions on which the single-excitation dynamics is built."},{"cited_title":"Subradiance-protected excitation transport,","cited_arxiv_id":null,"evidence_quote":"Supplies the dipole radiation kernel used to compute the interaction matrix and the far-field emission."}],"review_version":1}