{"id":"705aa19f-a325-4a3c-b100-c94310c7fee7","arxiv_id":"2607.20292","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":4,"one_line_summary":"Microwave-driven sympathetic cooling with same-isotope 43Ca+ ions cools a shared motional mode to n̄≈0.16 with 1.7(4)×10^-4 data-qubit error per cycle.","lead":"This paper demonstrates a way to cool the shared motion of two trapped ions using the same atomic species for data and coolant, driven by microwaves instead of additional lasers. The technique cools a two-ion mode to about 0.16 phonons while adding only 1.7×10^-4 error per cooling cycle to the data qubit, potentially simplifying trapped-ion quantum computers.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the disclosed coolant-depletion limitation affects long-term utility but not the central demonstration.","rationale":"The reader correctly identified coolant depletion as a limitation, and the paper's own supplement quantifies it clearly. However, the central claim is an experimental demonstration of cooling and error per cycle, not a demonstration of indefinite continuous operation. The depletion does not invalidate the measured n̄=0.16 or the 1.7×10⁻⁴ error. The stress-test found no more load-bearing concern: the thermometry, IRB methodology, error budget, and cooling-rate characterization are internally consistent and supported by the data shown. The reader's verdict of ACCEPT remains appropriate; the only adjustment would be to note the long-term replenishment requirement more prominently, but that is a scope note, not a correctness issue.","tokens_in":20804,"tokens_out":17650,"duration_ms":145053,"concrete_test":"Run an interleaved sequence of ~100 gates and cooling cycles on a two-ion crystal while maintaining n̄≈1, and measure the coolant population loss per cycle via state-selective fluorescence. Confirm the 3×10⁻³ loss rate and test whether shuttling a fresh coolant ion restores steady-state operation.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—cooling a two-ion mode to n̄≈0.16 and a data-qubit error of 1.7(4)×10⁻⁴ per cycle—is well supported by the presented thermometry and IRB data. The reader's concern about coolant depletion (Supplement S2B, ~3×10⁻³ loss per sequence when interleaved with gates) is a real, transparently disclosed limitation for continuous operation, but it does not undermine the correctness of the measured cooling performance or the per-cycle error benchmark. The authors explicitly state that the loss saturates in their measurement regime because the cooled population decreases, and they suggest replenishment strategies. The two-ion validation for shorter IRB sequences further supports the error claim. No internal inconsistency or unsupported assumption that would threaten the central experimental results was identified.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"Smith et al. demonstrate a same-isotope sympathetic cooling scheme for 43Ca+ ions in which both data and coolant ions remain in the ground-state hyperfine manifold and the coolant is addressed through microwave-driven resolved sideband cooling at a Zeeman-shifted transition. The shared in-plane radial rocking mode is cooled to n̄=0.16(2) quanta after 16 cycles, with a cooling rate of 0.27(1) quanta/cycle at n̄=1. Using interleaved randomized benchmarking, they measure a data-qubit error of 1.7(4)×10−4 per cooling cycle, compared with an idle error of 0.7(2)×10−4. Ac-Zeeman compensation tones and spin-echo pulses are used to protect the data qubit. The manuscript includes a detailed supplementary characterization, an error budget, and simulations of the repumping dynamics.","tokens_in":20947,"tokens_out":10672,"duration_ms":99661,"significance":"The result is timely and important: it removes the need for a second species or tightly focused lasers for sympathetic cooling and keeps all control in the microwave domain, which is attractive for integrated QCCD architectures. The experimental execution is careful: sideband thermometry with joint red/blue fits, IRB with parametric bootstrapping, two-ion cross-checks, and a component-level error budget. The authors also transparently disclose the coolant-population depletion in the repumping loop (Suppl. S2B), which is a real limitation for continuous interleaved operation but does not invalidate the measured per-cycle error. If the data hold, this is a useful step toward fully microwave-based sympathetic cooling.","major_comments":[],"minor_comments":[{"comment":"The conclusion states that maintaining n̄=1 would require ~50 cooling cycles per second and estimates a per-gate error contribution of ~1e-6. However, Suppl. S2B reports that when cooling is interleaved with gates, the coolant population is lost at ~3e-3 per sequence because the 'cooled' population no longer decreases, and no replenishment is implemented in this work. Please add an explicit caveat in the main text connecting this extrapolation to the need for replenishment (e.g., shuttling in fresh coolant ions), or temper the statement. This is a presentation issue and does not change the measured central claim.","section":"Conclusion & Suppl. S2B"},{"comment":"Minor typos: 'between between' (page 1), 'seperation' (page 1), 'processs' (page 4), 'occuring' (page 3), and 'n reproducible frequency shift' in the Fig. S4 caption (likely 'a reproducible').","section":"Throughout"},{"comment":"Reference [60] ('Atomic Physics, GitHub repository') is incomplete; please provide the repository URL, version/commit, and the appropriate citation format so that the simulation can be reproduced.","section":"Ref. [60]"},{"comment":"The main text states that the two-ion IRB validation used 'shorter gate sequences.' Please state explicitly how many two-ion sequences were acquired and the values/error bars of those points, so the reader can assess the consistency of the single-ion and two-ion error estimates.","section":"Fig. 3 / IRB"}],"recommendation":"minor_revision","confidential_remarks":"This is a strong experimental demonstration. The central measured claims—cooling to n̄≈0.16 and an error of 1.7(4)×10−4 per cooling cycle—are well supported. The main substantive caveat is that the conclusion's continuous-operation extrapolation should be reconciled with the coolant-depletion limitation disclosed in S2B; this can be fixed with text changes. The comparison with the recent Duke preprint is handled fairly. I recommend minor revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"John — this is a solid experimental paper and the central claim holds up. The authors cool a two-ion radial rocking mode to n̄ = 0.16(2) using a same-isotope scheme where the coolant transition is Zeeman-separated from the data qubit within the 4S1/2 manifold, and they measure a data-qubit error of 1.7(4)×10⁻⁴ per cooling cycle via IRB. That is the first time microwave-only, same-isotope sympathetic cooling has been shown to preserve qubit coherence, and it removes the need for a second species or tightly focused lasers. The experiment is well executed: sideband thermometry with joint red/blue fits, IRB with parametric bootstrapping, and a decomposed error budget showing the sideband pulse dominates (1.2×10⁻⁴), followed by transfer pulses (0.3×10⁻⁴), with laser pulses indistinguishable from idle. They are transparent about the IRB being mostly single-ion with two-ion validation for shorter sequences.\n\nThe main caveat is in Supplement S2B: the repumping loop is not closed. They estimate about 3×10⁻³ of the coolant population is lost per sequence when cooling is interleaved with gates to maintain n̄ ≈ 1, because the 'cooled' population no longer decreases. In their demonstration, the loss saturates because they just run many cooling cycles without gate heating. They do not implement coolant replenishment. So for continuous QCCD operation, someone would need to shuttle in fresh coolant ions or fix the repumping losses. That is a real engineering limitation, but it is disclosed, quantified, and doesn't invalidate the measured per-cycle error or the cooling rate.\n\nThe other soft spot is that the cooling rate is modest — 0.27(1) quanta/cycle, 290 quanta/s — and the cycle time is 920 μs. The authors acknowledge that cryogenic operation and MW power management would improve both rate and error. The fundamental scattering limit is estimated at 1.6×10⁻⁶ per cycle, so there is headroom.\n\nI checked the citation pattern; it covers the prior same-isotope work (Rohde, Sriarunothai, the Duke preprint) and the relevant laser-free logic references. No red flags.\n\nThis paper is for the trapped-ion QCCD community, especially groups using microwave-driven logic. It deserves a serious referee, not a desk reject. The main suggestion for revision would be to discuss more concretely how the coolant replenishment could be integrated, and perhaps to show thermometry data at higher initial n̄ or after interleaving with gates. But as a demonstration, it stands.","headline":"Solid experimental demonstration of microwave-driven same-isotope sympathetic cooling with careful benchmarking; the coolant-depletion caveat is real but disclosed and does not sink the central result.","tokens_in":21503,"tokens_out":2223,"would_cite":true,"duration_ms":21494,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.-a","37.10.Ty","37.10.Rs"],"model":"deepseek-v4-flash","headline":"Same-isotope microwave cooling brings ion motion to 0.16 quanta with a per-cycle data-qubit error of 1.7×10⁻⁴.","keywords":["sympathetic cooling","trapped ions","microwave control","Zeeman splitting","sideband cooling","same-isotope cooling","interleaved randomized benchmarking","43Ca+"],"falsifier":"Run the interleaved randomized benchmarking sequence for many more cooling cycles (e.g., 10⁴) while monitoring the coolant ion population. If the coolant population decays at the ~3×10⁻³ per sequence rate reported in Supplement S2B and the cooling rate correspondingly degrades, the scheme cannot sustain n̄ ≈ 1 without coolant replenishment.","tokens_in":20670,"feed_emoji":"🔬","tokens_out":4375,"duration_ms":36276,"temperature":0.7,"pith_summary":"This paper establishes that sympathetic cooling of a data qubit can be done with a coolant ion of the same species and isotope, using only global near-field microwaves, by separating the coolant and data transitions in frequency via a static magnetic field's Zeeman splitting. In ⁴³Ca⁺, pulsed resolved sideband cooling on a coolant transition lowers the shared in-plane radial rocking mode to an average phonon occupation of 0.16(2), with a cooling rate of 0.27(1) quanta per cycle at n̄ = 1. Interleaved randomized benchmarking puts the induced data-qubit error at 1.7(4)×10⁻⁴ per cooling cycle, of which 0.7(2)×10⁻⁴ is idle error. The authors attribute most of the residual error to thermal drifts in the microwave drive chain and estimate a fundamental photon-scattering floor near 1.6×10⁻⁶ per cycle. They also note, in the supplement, that coolant population is lost when cooling is interleaved with gates and that replenishment is not demonstrated.","feed_headline":"Microwaves cool ion motion to 0.16 quanta, no second species","feed_subtitle":"Same-isotope sympathetic cooling adds only 1×10⁻⁴ per cycle to data-qubit error, simplifying trapped-ion laser setups.","key_machinery":"Zeeman-frequency separation within the ground-state manifold, enabled by a homogeneous 28.8 mT static magnetic field. The global near-field microwave tone drives a resolved sideband on the coolant transition |4,2⟩→|3,2⟩ while the data clock transition |4,1⟩→|3,1⟩ is detuned by about 200 MHz. Simultaneous detuned compensation tones cancel ac Zeeman shifts on both coolant and data transitions, and spin-echo pulses applied between pairs of cooling cycles cancel residual shifts; SK1 composite pulses stabilize against amplitude noise from thermal transients.","core_discovery":"The paper's central claim is that sympathetic cooling of a trapped-ion data qubit can be driven by global near-field microwaves without a second species, by using the Zeeman splitting of a static magnetic field to isolate the coolant transition from the data qubit transition in frequency space. With ⁴³Ca⁺, the data qubit lives in clock states |4,1⟩ and |3,1⟩, while the coolant uses the |4,2⟩→|3,2⟩ transition; a resolved sideband on this coolant transition cools the shared motion to n̄ = 0.16(2). The measured data-qubit error per cooling cycle is 1.7(4)×10⁻⁴, benchmarked by interleaved randomized benchmarking, and the authors show that the error budget is dominated by technical ac Zeeman shif","pith_inferences":["Extension: A natural next step, not demonstrated here, is to integrate coolant replenishment—for example, shuttling a freshly prepared coolant ion from another zone—so that the scheme can run indefinitely when interleaved with gates; the paper's own supplement shows the coolant population drops by roughly 3×10⁻³ per sequence in that regime.","Extension: The compensation-tone approach suggests a general strategy: any pair of transitions driven by the same microwave chain can have common-mode ac Zeeman drifts rejected, which may carry over to other multi-tone microwave control schemes in trapped-ion processors.","Extension: A testable improvement would be to use a coolant transition farther from the data qubit in frequency (for example, |4,−3⟩→|3,−3⟩ in ⁴³Ca⁺) or a different species with a larger Zeeman manifold; the paper notes the trade-off but does not implement it.","Extension: Since both cooling and entangling logic can be microwave-driven, this scheme points toward a largely laser-free trapped-ion processor, though the present demonstration only benchmarks single-qubit gate sequences and not entangling gates."],"forward_implications":["Same-isotope sympathetic cooling removes the need for a second ion species and its extra laser wavelengths; the same microwave control used for logic can drive cooling.","The measured data-qubit error of 1.7(4)×10⁻⁴ per cooling cycle, with idle error 0.7(2)×10⁻⁴, implies a net cooling-induced error of about 1.0×10⁻⁴ per cycle.","Maintaining n̄ ≈ 1 would require roughly 50 cooling cycles per second; at that occupation the scheme's contribution to a two-qubit gate error would be around 10⁻⁶, below current entangling-gate error rates.","Cryogenic operation and microwave power management are identified as routes to lower the technically dominated error, with a fundamental photon-scattering limit about two orders of magnitude below the measured value.","The scheme is claimed to be applicable to any ion species with metastable states and at least four states in the ground-state manifold."],"fun_headline_variants":["Same-species ion cooling via microwaves: 0.16 quanta, 1.7e-4 error","Microwave-only sympathetic cooling for trapped ions to n̄=0.16","No second species: microwave cooling of ion motion, error 1.7e-4","Simplify ion traps: same-isotope microwave cooling to 0.16","Ion sympathetic cooling with microwaves: same species, 0.16 n̄"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The scheme's long-term usefulness depends on replenishing coolant ions, because when cooling is interleaved with gates each sequence loses about 3×10⁻³ of the coolant population, and the paper does not demonstrate such replenishment.","fun_headline_variants_meta":{"raw":{"variants":["Same-species ion cooling via microwaves: 0.16 quanta, 1.7e-4 error","Microwave-only sympathetic cooling for trapped ions to n̄=0.16","No second species: microwave cooling of ion motion, error 1.7e-4","Simplify ion traps: same-isotope microwave cooling to 0.16","Ion sympathetic cooling with microwaves: same species, 0.16 n̄"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000452,"raw_usage":{"total_tokens":2077,"prompt_tokens":677,"completion_tokens":1400,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":421,"completion_tokens_details":{"reasoning_tokens":1285}},"tokens_in":421,"tokens_out":1400,"duration_ms":10936,"temperature":1.0,"reasoning_tokens":1285,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T10:13:52.219049+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the interleaved randomized benchmarking sequence for many more cooling cycles (e.g., 10⁴) while monitoring the coolant ion population. If the coolant population decays at the ~3×10⁻³ per sequence rate reported in Supplement S2B and the cooling rate correspondingly degrades, the scheme cannot sustain n̄ ≈ 1 without coolant replenishment.","supporting_citations":[],"review_version":1}