{"id":"d64917fe-fb68-42d6-9703-3fd19cc7b116","arxiv_id":"2607.25409","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A glass-walled stellarator using only six identical circular coils confines low-temperature argon, neon, and helium plasmas inside computed vacuum surfaces with ι≈0.3.","lead":"Polaris is a new small stellarator built at the Swiss Plasma Center: six identical circular copper coils sit inside a mostly glass vacuum vessel and hold a low-temperature plasma in a 3D toroidal shape. The device is designed to swap coil sets easily, making it a flexible testbed for stellarator edge physics and a visual outreach platform.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"As-built coil geometry is not directly verified: a single on-axis B0 measurement and visual plasma shape are insufficient to establish the claimed closed flux surfaces and rotational transform.","rationale":"The reader's weakest assumption is exactly the most load-bearing: the paper's experimental results are interpreted as demonstrating confinement in a designed stellarator field, but the field itself is not measured. The B0 check is a single scalar magnitude, not a topology verification; the plasma shape is qualitative and could be produced by local mirror effects. The paper's own future-work statement (fluorescent rod) confirms the absence of this measurement. This concern is not an ad hominem or an appeal to consensus; it is an internal gap between what is claimed (closed surfaces, rotational transform) and what is verified (plasma visible, B0 magnitude). The perturbation experiment provides some supporting evidence that the optimized geometry matters, but it does not isolate the role of rotational transform as opposed to field strength or mirror ratio. A conditional verdict is appropriate because the issue is addressable with a relatively simple diagnostic; there is no indication of fraud, just missing evidence. The recommendation is UNCHANGED because the reader already set CONDITIONAL and this stress-test reinforces rather than alters that judgement.","tokens_in":14829,"tokens_out":5099,"duration_ms":58364,"concrete_test":"Perform vacuum flux-surface mapping using a fluorescent rod (or an electron beam) at the nominal coil positions and current. Insert a probe radially at multiple toroidal angles, trace the surfaces, and compare with Poincaré sections from field-line tracing of the as-built coil geometry (measured by photogrammetry). If nested, island-free surfaces with rotational transform matching 0.27<ι<0.31 are found over a volume ~0.05 m³, the concern is resolved. If islands appear or surfaces deviate beyond the stated tolerance, the central claim that Polaris confining plasma by closed surfaces is not credible.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that Polaris produces vacuum magnetic surfaces with substantial rotational transform and that the plasma is confined by this 3D field—rests on the as-built coils realizing the optimized SIMSOPT geometry within the δx<1 cm tolerance. The paper reports no coil metrology or direct magnetic-surface mapping. The only checks are the on-axis field B0≈120 G at the coil center (Section IV, B), which depends mainly on coil radius and current and is insensitive to misalignment, and the visible match of the plasma to the computed last closed surface (Section IV, A). A bright plasma in a collisional, high-neutral-pressure regime can appear to follow the toroidal field and local mirror structure even without closed flux surfaces; the strong mirror ratio Bmax/Bmin≈10 (Section II) could locally confine plasma between adjacent coils. The perturbation experiment (coil 2A rotated 180°) degrades τ2 by only ~30%, and the plasma still fills the torus, showing that even a non-stellarator field confines plasma. The paper itself lists fluorescent-rod magnetic-surface mapping as future work (Section V), acknowledging this missing verification. Without quantitative confirmation of the vacuum magnetic topology, the evidence for rotational transform and closed surfaces is entirely theoretical, and the experimental 'confinement' might be due to mirror effects rather than the optimized 3D field.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper describes the design, construction, and first plasma experiments of Polaris, a small stellarator at the Swiss Plasma Center with six identical circular modular coils arranged inside a glass vacuum vessel. The coil geometry was obtained by single-stage optimization with SIMSOPT (Ref. 17), giving vacuum magnetic surfaces with rotational transform 0.27<iota<0.31 over about 0.05 m^3 and a nominal robustness to coil displacements below 1 cm. The device uses a 13.56 MHz ICP antenna for plasma production, water-cooled copper coils, and full optical access. First plasmas in argon, neon, and helium are reported, with n_e~1e16-1e17 m^-3, T_e~2-6 eV, a dominant 109 kHz fluctuation, and a two-timescale decay of ion saturation current after RF switch-off, with the slow time tau2 about 1.17 ms at 5e-3 mbar. A perturbation experiment, in which one coil is rotated by 180 degrees, shows a ~30% reduction of tau2 and changes in density profiles, interpreted as evidence that the optimized configuration contributes to confinement.","tokens_in":15171,"tokens_out":6790,"duration_ms":71387,"significance":"If the claims are sustained, Polaris is a valuable new university-scale stellarator testbed. Its key strengths are the demonstration of a very simple modular coil set (six identical circular coils) producing substantial rotational transform, a large glass vacuum vessel with full optical access, a flexible coil-inside-vessel concept that allows rapid configuration changes, and a first set of plasma diagnostics including imaging, Langmuir probe profiles, fluctuation spectra, and a coil-perturbation study. The use of standard computational tools (SIMSOPT, VMEC, booz_xform) and the FEA validation of window deformations are positive features. The main limitation is that the experimental verification of the vacuum magnetic topology is indirect, as the as-built coil positions are not measured and no magnetic surface mapping is performed. This limits the strength of the claim that the plasma is confined by the optimized closed flux surfaces rather than by mirror/antenna effects.","major_comments":[{"comment":"The central experimental claim—that the plasma bulk is confined by the optimized closed flux surfaces with rotational transform—is not directly verified. The only quantitative magnetic check is the on-axis field B0 ~ IB/1.25 = 120 G at the coil center (Sec. IV.B), which depends mainly on coil radius and current and is insensitive to the five optimized degrees of freedom. The visible match of the luminous plasma to the computed last closed surface (Sec. IV.A) is suggestive but, in a low-temperature, high-neutral-pressure plasma with Bmax/Bmin ~ 10 (Sec. II), a mirror-dominated discharge can also fill the torus and follow the field shape. The perturbation experiment (Sec. IV.E) shows that even with coil 2A rotated 180 degrees the plasma still fills the torus and tau2 is reduced by only about 30%, so toroidal propagation and slow decay do not by themselves imply closed flux surfaces. Sectio","section":"IV.A, IV.B, V"},{"comment":"The 'global confinement time' tau2 is inferred from a two-exponential fit to the ion-saturation decay at a single probe position (r=0, toroidal location 5b). Because n_e varies by 60-75% between toroidal locations 4e and 5b (Fig. 15) and the mirror ratio is large, a local decay time need not equal the global particle or energy confinement time; it could reflect local parallel redistribution or losses. In addition, the attribution of tau1 to electron-neutral radiative losses is plausible but is not directly supported by time-resolved Te or radiation measurements. Please rename this quantity a 'local ion-saturation decay time' and, if the global statement is retained, support it with measurements at several toroidal/poloidal positions or with a model that accounts for the mirror geometry.","section":"IV.C"}],"minor_comments":[{"comment":"The text says surfaces are 'well preserved for perturbations up to delta_x < 1 cm', but the delta_x = 1 cm case in Fig. 4 shows a clear 1/3 island. Define the quantitative criterion for 'well preserved' (e.g., island width, fraction of destroyed flux, or maximum acceptable deviation) and clarify that the bound is strict.","section":"II, Fig. 4"},{"comment":"The mean n_e and T_e profiles in Fig. 13 are shown without error bars. Since the fluctuations of V_f are reported with standard deviations, it would be helpful to also give shot-to-shot or statistical uncertainties for n_e and T_e, especially because the inset claims a linear dependence on ionization energy.","section":"IV.B, Fig. 13"},{"comment":"The caption of Fig. 15 says 'toroidal location 4e and 4b', while the text refers to 4e and 5b. Also, the text describes the modified-configuration profiles as 'purple curves' while the caption says 'black curves'; please reconcile.","section":"IV.D, Fig. 15"},{"comment":"In the confinement-time paragraph, the low pressure is given as 0.15e-3 mbar, which appears to be a typo for 1.5e-3 mbar. Please correct.","section":"IV.E"},{"comment":"The engineering sections are informative, but a few statements would benefit from clarification: the 'total current in each coil' in the abstract (~5 kA) should be explicitly defined as ampere-turns (16 turns times coil current), and the window deformation test results (experimental values ~30% lower than simulated) should be reported with uncertainties.","section":"III.A/III.B"},{"comment":"For a device paper, consider making the optimized coil geometry and the raw plasma profiles available in a public repository, rather than only 'upon reasonable request', to improve reproducibility.","section":"Data Availability"}],"recommendation":"major_revision","confidential_remarks":"This is a well-executed and clearly written device paper. The stress-test concern about the missing as-built magnetic verification is real and load-bearing: the experimental claim of confinement by closed flux surfaces rests on indirect evidence. I would not reject, because the design, engineering, and first-plasma dataset are valuable and the limitation is explicitly acknowledged in Section V. The revision should either add coil metrology / magnetic-surface mapping or carefully tone down the confinement claims, and should correct the local-vs-global interpretation of tau2. With those changes, the paper could be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Polaris is a genuine step forward for university-scale stellarator experiments. The new thing isn't the coil optimization — that's from Jorge et al. — it's the actual machine: a glass vacuum vessel that allows full optical access, coils inside the vessel, and a flexible layout that makes swapping configurations easy. The engineering is thoughtful, and the first plasma data, while preliminary, show a plasma that roughly follows the computed last closed surface, with densities and temperatures in the expected range. The perturbation experiment (rotating one coil) degrades confinement by ~30%, which is a useful sanity check that the field geometry matters.\n\nThe soft spots are real but not fatal. The paper never directly verifies the as-built vacuum magnetic topology. There is no coil metrology and no fluorescent-rod surface mapping; the evidence is a single on-axis B0 measurement and the visible plasma shape. That shape could be affected by the strong mirror ratio and by the antenna location, so it isn't conclusive proof of closed flux surfaces. The authors acknowledge this and list mapping as future work, which is honest. The confinement time is a single-point ion-saturation decay with a two-exponential fit; the fast phase is attributed to radiation, the slow phase to transport, but that interpretation is plausible rather than proven. Profiles lack error bars. Data and code aren't shared, which is a limitation for a device paper.\n\nThe central claim — that a low-temperature plasma can be produced and roughly confined in the field of six identical circular coils — holds up. The paper does not claim more than that, and the conclusions are appropriately cautious. The stress-test concern about mirror confinement is worth taking seriously, but I don't think it sinks the paper. Even if some of the visible confinement is due to mirrors, the device still works as a flexible testbed, and the degradation in the perturbed configuration shows the optimization has some effect.\n\nWho is this for? Experimentalists building small stellarators, people interested in edge-relevant turbulence studies, and those who want a visual demonstration of 3D confinement. It deserves a serious referee. The main requests would be: add error bars, report coil positions or at least a tolerance verification, and share the SIMSOPT input. Those are standard for this kind of paper.\n\nI'd send it to peer review. It's not a high-impact physics result, but it's a useful experimental contribution that is honest about its own limitations.","headline":"A well-executed proof-of-principle for simple-coil stellarators, with honest first plasma results; the as-built magnetic topology isn't directly verified, but the paper doesn't oversell it.","tokens_in":15695,"tokens_out":2174,"would_cite":true,"duration_ms":23153,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["52.55.Hc"],"model":"deepseek-v4-flash","headline":"Polaris demonstrates that six identical, circular, water-cooled coils—optimized only in position and tilt—can generate a sizable volume of closed magnetic surfaces with rotational transform near 0.3 and confine a low-temperature plasma for","keywords":["stellarator","simple modular coils","rotational transform","vacuum magnetic surfaces","low-temperature plasma","RF inductively coupled plasma","coil error tolerance","plasma confinement time"],"falsifier":"Measure the actual coil positions (for example by photogrammetry or coordinate measurement) and map the vacuum magnetic surfaces directly (for example with an electron beam or fluorescent rod). If any coil is displaced more than about 1 cm, or if the measured last closed surface shows the 1/3 island chain predicted for that displacement, the claimed robustness and the interpretation of 'confinement' as magnetic-surface confinement would need revision.","tokens_in":14767,"feed_emoji":"🧲","tokens_out":6313,"duration_ms":59300,"temperature":0.7,"pith_summary":"Polaris is a small glass-walled stellarator built to test whether a plasma can be confined with the simplest possible coil set: six identical, circular, water-cooled coils arranged toroidally and optimized only by position and orientation. The paper reports that this configuration creates a 0.05 cubic meter volume of closed vacuum magnetic surfaces with rotational transform 0.27–0.31, and that the surfaces survive random coil displacements up to about 1 centimeter. First plasmas—argon, neon, and helium at 2–6 eV and 1e16–1e17 m^-3—light up along the predicted last closed surface, and the measured global confinement time is about a millisecond. Deliberately rotating one coil by 180 degrees cuts the confinement time by about 30 percent, evidence that the coil optimization itself contributes to confinement. The device's transparent vessel and interchangeable base plates are meant to make stellarator physics visibly accessible and to open a testbed for edge-relevant turbulence, radiation, and neutral-plasma interaction.","feed_headline":"Six identical circular coils confine plasma in Polaris","feed_subtitle":"A glass-walled testbed shows rotational transform near 0.3 and centimeter-scale coil tolerance, opening edge physics to direct view.","key_machinery":"The enabling design element is the optimized arrangement of six identical circular coils (25 cm diameter, 16 turns each) inside a vessel, leaving 5 free degrees of freedom under 3-fold periodicity and stellarator symmetry. The coils are first optimized as filaments and then evaluated as finite-width bundles; the rotational transform is produced by the torsion of the magnetic axis rather than by coil helicity. A second load-bearing component is the all-glass vacuum vessel, which gives 360-degree optical access and fast interchange of coil sets via replaceable base plates; the in-vessel RF antenna with a Faraday shield and ceramic casing ignites the plasma inductively.","core_discovery":"The paper's central claim is that a stellarator with substantial rotational transform and practical error tolerance can be built from six identical planar circular coils, positioned and tilted by single-stage optimization. Rotational transform arises from the integrated torsion of the magnetic axis, not from coil helicity: iota approximately equals (1/2pi) times the line integral of torsion minus N. Numerically, the configuration yields 0.05 cubic meters of closed vacuum flux surfaces with 0.27 < iota < 0.31, and surfaces survive random coil displacements below 1 centimeter. Experimentally, the plasma glow tracks the computed last closed surface, densities and temperatures are typical of a c","pith_inferences":["If the centimeter-scale tolerance extrapolates to larger devices, simple-coil stellarators might be built without the precision metrology typical of big optimized machines; a natural test is deliberately scanning coil displacement past 1 cm and mapping the resulting island widths.","The optical transparency invites direct validation of field-line tracing: a fluorescent rod inserted along predicted surfaces, as the paper suggests, would confirm the nested-surface picture that currently rests on one on-axis field reading and the plasma glow.","The 109 kHz fluctuation, once its poloidal wavenumber is measured, could be used to test whether drift or interchange scaling laws hold in a stellarator edge analog, giving low-cost data for turbulence transport models.","Because the vessel permits rapid coil swaps, Polaris could systematically scan the simple-coil configuration space—varying tilt angles and positions—to produce an empirical map of which nearby arrangements confine best, complementing numerical optimization."],"forward_implications":["Simple identical circular coils can serve as a proof-of-principle for low-complexity stellarator optimization, potentially reducing coil manufacturing costs.","A 1-cm coil positioning tolerance, several percent of the major radius, is much looser than typical optimized stellarator designs and simplifies assembly.","The measured two-phase plasma decay after RF turn-off separates radiation losses (fast, about 30 microseconds) from cross-field transport (slow, about 1 millisecond) in this collisional regime.","Rotating one coil by 180 degrees degrades the global confinement time by about 30 percent, showing the optimized field measurably improves confinement even in a low-temperature, neutral-dominated plasma.","Polaris provides a platform for studying drift or interchange instabilities and neutral-dominated edge physics in a three-dimensional magnetic geometry."],"fun_headline_variants":["Six circular coils make a flexible stellarator","Glass fish-tank tests stellarator edge physics","Planar coils generate twist without helicity","Stellarator fish-tank: optical access to plasma","Rotational transform from six planar coils"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The whole confinement story rests on the as-built coils matching the optimized geometry within the about-1-cm tolerance, but the paper does not report coil metrology or direct magnetic-surface mapping—only a single on-axis field measurement and the apparent match of the luminous plasma to the computed boundary.","fun_headline_variants_meta":{"raw":{"variants":["Six circular coils make a flexible stellarator","Glass fish-tank tests stellarator edge physics","Planar coils generate twist without helicity","Stellarator fish-tank: optical access to plasma","Rotational transform from six planar coils"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00046,"raw_usage":{"total_tokens":2183,"prompt_tokens":831,"completion_tokens":1352,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":575,"completion_tokens_details":{"reasoning_tokens":1297}},"tokens_in":575,"tokens_out":1352,"duration_ms":11295,"temperature":1.0,"reasoning_tokens":1297,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T02:31:02.534517+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the actual coil positions (for example by photogrammetry or coordinate measurement) and map the vacuum magnetic surfaces directly (for example with an electron beam or fluorescent rod). If any coil is displaced more than about 1 cm, or if the measured last closed surface shows the 1/3 island chain predicted for that displacement, the claimed robustness and the interpretation of 'confinement' as magnetic-surface confinement would need revision.","supporting_citations":[],"review_version":1}