{"id":"f1e6a6ac-22d8-42b5-b4ab-96b04c0a3a52","arxiv_id":"2506.03883","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"In ASDEX Upgrade helium plasmas, He2+ dominates tungsten divertor erosion, and the chosen sheath boundary condition can change the predicted tungsten source by an order of magnitude.","lead":"This paper models how helium plasma erodes tungsten in a tokamak divertor, using two simulation codes and a simple analytical formula. It finds that doubly charged helium ions dominate the erosion and that code choices for the plasma boundary strongly change the predicted tungsten source.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"T_i is not directly validated (Sec. 4.3), and the paper's own sensitivity scan shows factor-2 to -3 changes in T_i/T_e alter net erosion by order(s) of magnitude; this uncertainty could flip the He2+-dominance crossover (Sec. 3) and change the reported W-source boundary-condition factor.","rationale":"The reader's weakest assumption (T_i not validated, constant extrapolation) identifies the same broad area as this stress test. I focused specifically on T_i because the paper itself shows that net erosion is extremely sensitive to it (Figure 6) and because the analytical crossover at ~65 eV makes the He2+-dominance claim potentially fragile under plausible T_i variation. This is not an internal inconsistency; the authors are transparent about the limitations. The question is whether the central claims—He2+ dominance even at equal ion densities, and the factor-20 overestimate from the collective-velocity sheath boundary condition—survive within the stated uncertainty envelope. Since the reader already assigned a CONDITIONAL verdict on the basis of insufficient experimental validation of key inputs, this stress test does not change that verdict but sharpens the specific condition: demonstrate that the relative results are stable when T_i is varied over the range used in the paper's own sensitivity study. I therefore recommend UNCHANGED rather than a more severe verdict. The paper's strengths include a well-documented workflow, use of two independent codes, explicit comparison of boundary conditions, and honest discussion of the experimental mismatch; these support a conditional acceptance pending the T_i robustness check.","tokens_in":17767,"tokens_out":12074,"duration_ms":112656,"concrete_test":"Re-run the ERO2.0 baseline with the ion temperature channel multiplied by the factors used in the paper's own Figure 6 (1.5, 2, 3) and recompute (a) the ratio of He2+-induced to He+-induced gross erosion in the outer strike-point region and (b) the W-source comparison from Table 1 (SOLPS-ITER with both boundary conditions vs ERO2.0). If the He2+/He+ gross-erosion ratio remains >1 and the collective-BC source remains roughly 20x the species-dependent one for all multipliers, the qualitative claims survive; if the ratio crosses 1 or the factor drops significantly, the claims are not robust to the stated T_i uncertainty.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing assumption is that the SOLPS-ITER background, specifically T_i (and the extrapolated near-wall profiles), is accurate enough for quantitative erosion predictions. Section 4.3 states that T_i is not directly validated, and Figure 6 shows that multiplying T_e or T_i by 1.5-3 increases net erosion by more than an order of magnitude; agreement with the measured ~50-100 nm eroded thickness requires T_e x3 or combined T_e x2, T_i x2 plus O6+ at several percent. The central relative claims are not independent of this uncertainty. The analytical model in Section 3 predicts a crossover at T~65 eV (normal incidence) below which He2+ dominates despite carrying half the particle flux; the ERO2.0 result in Section 4.2 depends on the simulated T_e, T_i staying below that crossover. If the true T_i is near the upper end of the uncertainty band, sputtering yields for He+ and He2+ converge and the 'He2+ dominates even at 50/50 density' claim may reverse. Similarly, the W-source comparison in Table 1 (collective vs species-dependent BC, 90.08 vs 5.08 x 10^19 W/s) is computed at these low temperatures; the W self-sputtering yield is steeply energy-dependent, so a hotter background could materially change the 20x factor. Without an independent T_i measurement or a documented T_i sensitivity scan for the relative quantities, the central conclusions are conditional on an unvalidated input.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript models tungsten erosion and scrape-off-layer transport in the L-mode helium discharge #36687 at ASDEX Upgrade, combining a simplified analytical sputtering model, SOLPS-ITER plasma background and impurity simulations, and ERO2.0 kinetic impurity tracing. The SOLPS-ITER background is benchmarked against outboard midplane and outer divertor diagnostics. The authors report that the divertor ion flux near the strike point is approximately 50% He+ and 50% He2+, with He2+ dominating farther into the scrape-off layer. An analytical model predicts that for Te=Ti below about 65 eV, erosion increases with He2+ fraction; ERO2.0 with the SOLPS background confirms that the mixed-fraction case is closer to the full-He2+ case than to the full-He+ case. Net erosion predicted by ERO2.0 is more than an order of magnitude below the 50–100 nm measured post-mortem, and agreement is recovered only by combining elevated temperatures (factor 2) with a fitted O6+ concentration of a few percent. In the final section, SOLPS-ITER simulations with tungsten show that the choice of sheath boundary condition (collective fluid velocity vs species-dependent sound speed) changes the gross tungsten source by about a factor 20, mainly through self-sputtering, and that ERO2.0 transports more tungsten towards the X-point, an effect that can be partly compensated by reducing the anomalous diffusivity D_W.","tokens_in":18131,"tokens_out":6683,"duration_ms":62579,"significance":"The paper addresses a relevant gap in quantitative plasma-wall interaction modelling for helium plasmas, where the ion charge state is not directly measurable. If the results hold, the analytical crossover criterion and the ERO2.0/SOLPS comparison are useful contributions to the community. Strengths include the benchmarking of SOLPS-ITER against outboard midplane and outer divertor data, an analytical model that is parameter-free in the sense of depending only on standard sputtering data and the SOLPS-derived saturation current, an explicit sensitivity analysis over Te, Ti, incidence angle, and oxygen concentration, and a clear quantification of the effect of sheath boundary conditions on the tungsten source in Table 1. The main caveat is that the apparent agreement with experimental erosion in Fig. 8 is obtained by fitting the unmeasured oxygen concentration after applying temperature multipliers, and the ion temperature field is not independently validated; the paper would be strengthened by presenting the He2+ versus He+ dominance and the boundary-condition factor as functions of the Ti/Te uncertainty.","major_comments":[{"comment":"The quantitative comparison with post-mortem erosion is not a prediction: the baseline ERO2.0 result (a few nm) is more than an order of magnitude below the measured 50–100 nm, and the agreement in Fig. 8 is obtained only after applying an ad hoc factor 2 to both Te and Ti and treating the O6+ concentration as a free parameter in the range 0.1–3%. Because the final profile is produced by this fitted combination, the statement in the conclusions that the model could be considered general overreaches. Please state explicitly that the Fig. 8 agreement is a fit, and provide the resulting uncertainty band on the inferred oxygen concentration and on the predicted eroded depth.","section":"§4.3, Figs. 6–8"},{"comment":"The central claim that He2+ dominates even where the two helium ion species are roughly equal in density depends on the simulated Te and Ti remaining below the ~65 eV crossover of the analytical model. The paper states in Sec. 4.3 that Ti is not directly validated, and Fig. 6 shows that a factor 2–3 increase in Ti changes net erosion by orders of magnitude. Under a hotter Ti, the sputtering yields for He+ and He2+ converge, so the relative contribution could reverse. Please add a sensitivity scan of the ERO2.0 He2+ versus He+ erosion ratio over the plausible Ti range, or provide an independent constraint on Ti near the strike point.","section":"§4.3 and §3"},{"comment":"The factor of about 20 between the collective and species-dependent sheath boundary conditions (90.08 vs 5.08×10^19 W/s) is computed at the baseline low-temperature background. Since the difference is attributed mainly to W self-sputtering, which is steeply energy-dependent, the factor may itself be sensitive to the same Te/Ti uncertainty identified in Sec. 4.3. The manuscript should either show the ratio as a function of background temperature or explicitly qualify the 20-fold statement as conditional on the SOLPS background.","section":"§5.1–5.2, Table 1"}],"minor_comments":[{"comment":"The 'notch around 0.2 m' is described as a geometry artifact; it should be masked or removed from the published figure to avoid confusion.","section":"Fig. 6"},{"comment":"The color scale and level curves are difficult to read because the panel labels repeat; consider a single shared color bar and larger labels.","section":"Fig. 3"},{"comment":"The text says that |V_sh| is often approximated as 3Te, but the full logarithmic expression is also given; please define the symbols me, mp, and A, and clarify when the 3Te approximation is used.","section":"Eq. (2)"},{"comment":"The 'jett' bundling model is not defined; a one-sentence description of the bundling method would help readers understand the 24 bundled tungsten species.","section":"§5.1"},{"comment":"The incidence-angle distribution obtained from the sheath tracing module is described but not shown; a plot or a quantitative statement of the distribution width would support the choice of 80°.","section":"§4.2"},{"comment":"The analytical model values in Table 1 should be compared only against the ERO2.0 quantity Γ_TOT − Γ_SELF, as the text correctly notes that self-sputtering is absent from the analytical model; the table caption should state this explicitly.","section":"§5.2, Table 1"},{"comment":"The conclusion that lowering D_W in ERO2.0 reproduces the SOLPS W density pattern is based on W density, while the abstract mentions 'W influx in core'; please clarify whether the comparison is on density or on flux.","section":"§5.3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal and the modelling is generally careful. The main weakness is that the experimental-erosion match is obtained by fitting two uncertain parameters (temperature multipliers and O6+ concentration), and the paper should be revised to present this as an inference with uncertainty rather than as a validation. I would not reject; the code-comparison results, especially the boundary-condition effect in Table 1, are useful for the SOLPS and ERO2.0 communities. No concerns about citation practice; earlier work by the same groups is cited appropriately."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about this paper. First, the new analytical erosion model with He+/He2+ fraction (Eq. 3) is a genuinely useful tool, and the finding that the SOLPS-ITER collective-velocity sheath boundary condition inflates W self-sputtering by ~20x compared to species-dependent sound speeds is practically important for anyone running impurity simulations. Second, the paper is honest that matching the measured erosion requires combining a factor-2 temperature increase with a fitted oxygen concentration — so the quantitative agreement is a fit, not a prediction.\n\nWhat the paper does well: the SOLPS-ITER background is benchmarked against outboard midplane and outer divertor data; the 50-50 He+/He2+ split near the strike point is a new simulation result; the sheath boundary-condition comparison (Table 1) is clearly presented; and the ERO2.0 vs SOLPS-ITER transport comparison is a thoughtful code-to-code exercise. The authors also state plainly that they underpredict erosion by more than an order of magnitude and discuss redeposition, sputtering-yield, and geometry uncertainties. That transparency counts for something.\n\nThe soft spots are real but not disqualifying. T_i is not directly validated, and the paper's own scans show that factor-2 to -3 changes in T_e or T_i alter net erosion by orders of magnitude. The stress-test note is correct: the He2+-dominance crossover at ~65 eV and the 20x boundary-condition factor are computed at low temperatures, and a hotter background could erode those relative conclusions. The final match with experiment is a multi-parameter fit (T_e x2, T_i x2, O6+ at a few percent), so the reader should not walk away thinking the model predicts erosion from first principles. There is also no shipped code or input deck, and the inner divertor and W transport results lack direct experimental validation.\n\nWho is this for? People doing edge plasma modeling with SOLPS-ITER or ERO2.0, especially anyone setting impurity boundary conditions or interpreting He-plasma experiments. It deserves a serious referee: the analytical model and the boundary-condition finding are worth publishing, and the limitations are acknowledged rather than hidden. I would send it to review, with a referee request for a T_i sensitivity scan of the relative quantities (He2+ dominance, the 20x factor) and a clearer separation between predicted and fitted results.","headline":"Solid applied modeling paper with a genuinely useful analytical erosion model and an important SOLPS boundary-condition finding, but the quantitative erosion match is a multi-parameter fit and T_i uncertainty could shift the central relative claims.","tokens_in":18745,"tokens_out":2183,"would_cite":true,"duration_ms":23320,"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":"In helium tokamak plasma, He2+ governs tungsten divertor erosion, and a standard fluid sheath boundary condition inflates the tungsten source about 20-fold.","keywords":["helium plasma","tungsten erosion","plasma-wall interaction","scrape-off layer","divertor","SOLPS-ITER","ERO2.0","sheath boundary condition"],"falsifier":"Take a repeat discharge with the same settings as #36687, measure the divertor ion temperature and O6+ density simultaneously, and compare ERO2.0's predicted erosion thickness against post-mortem marker measurements: if measured O6+ is below about 1% and erosion is still 50-100 nm, the model chain is missing something; if O6+ is a few percent and erosion matches, the impurity explanation is confirmed.","tokens_in":17584,"feed_emoji":"⚛️","tokens_out":6415,"duration_ms":62679,"temperature":0.7,"pith_summary":"The paper sets out to identify which helium ion charge state controls tungsten erosion at the divertor of a tokamak run in L-mode (low-confinement mode) helium plasma, and to test whether fluid and kinetic models agree on the resulting tungsten source and transport. It finds that although He+ and He2+ arrive in roughly equal numbers near the strike point, He2+ dominates the erosion because its double charge raises the energy gained in the Debye sheath and the local temperature is below the ~65 eV crossover of the sputtering curve. It also shows that the choice of sheath boundary condition in the fluid code matters enormously: imposing one collective outflow velocity for all ions instead of species-specific sound speeds raises the predicted tungsten source by a factor of about 20, almost entirely through artificial tungsten self-sputtering. Finally, it concludes that matching the experimentally measured erosion depth requires the presence of a few percent of highly charged light impurities, with oxygen as a proxy.","feed_headline":"He2+ drives tungsten erosion in helium tokamak divertors","feed_subtitle":"The doubly charged ion wins even at 50-50 mix; a standard fluid boundary condition inflates tungsten sources 20-fold.","key_machinery":"The argument is carried by two quantitative objects. The first is the He2+ fraction $f_{\\mathrm{He}^{2+}}$ and the sheath impact energy formula $E_{\\mathrm{wall}} = 2T_i + Ze|V_{\\mathrm{sh}}| \\sim 2T_i + 3ZeT_e$, combined with He-W sputtering yields: because the yield rises steeply with energy below about 65 eV, the doubly charged ion's extra sheath acceleration outweighs its halved particle flux, so erosion increases with $f_{\\mathrm{He}^{2+}}$. The second is the magnetic pre-sheath velocity boundary condition in the fluid code, comparing the collective fluid velocity with the single-species ion sound speed; switching between them changes the predicted tungsten source by a factor of about 20 through self-sputtering, with the species-dependent condition matching the kinetic Monte-Carlo result.","core_discovery":"On the paper's own terms, the central discovery is that divertor tungsten erosion in attached L-mode helium plasmas is carried overwhelmingly by He2+ ions, even where the He+ and He2+ fluxes are comparable, and that standard fluid simulations with a collective sheath velocity boundary condition overestimate the tungsten source by roughly a factor of 20 relative to kinetic modelling. The overestimate is traced to tungsten self-sputtering: when W ions are forced to leave at the collective helium sound speed, they strike the targets with inflated energy, whereas species-specific Bohm speeds and kinetic tracing give much lower self-sputtering. The same kinetic model, however, underpredicts the net erosion measured on exposed tungsten samples by more than an order of magnitude unless the plasma contains a few percent of highly charged oxygen, suggesting that extrinsic impurities rather than helium sputtering are the dominant measured erosion driver.","pith_inferences":["The temperature crossover of the erosion ranking is a general consequence of the sputtering-yield curve shape, so in hotter divertor conditions (above about 65 eV for He on W at normal incidence) the ranking could invert and He+ might dominate; this is an extrapolation of the analytical model, not a claim the paper makes for other devices.","The sheath boundary-condition sensitivity likely applies to any heavy impurity in fluid edge codes, including tungsten in deuterium-tritium plasmas, not only helium; a re-run with hydrogenic main ions would be a direct test.","Coupling kinetic-code tungsten sources into the fluid code with species-dependent Bohm speed would be a practical recipe for divertor impurity studies while avoiding the artificial self-sputtering inflation.","Spectroscopic measurement of the charge-state distribution of tungsten ions arriving at the targets could discriminate between the fluid and kinetic transport models, since the fluid model predicts a distribution peaked at higher charge states than the kinetic one."],"forward_implications":["Erosion estimates for helium plasmas must track He+ and He2+ separately; assuming a fully singly charged plasma underestimates the tungsten source by more than an order of magnitude in the conditions studied.","Fluid impurity simulations that default to perfect entrainment at the target can overstate heavy-impurity self-sputtering by about 20 times; the single-species sound-speed condition is the closer match to kinetic results.","Net erosion of marker samples in the strike-point region can only be reproduced when a few percent of high-charge oxygen is included, so measured erosion in such discharges points to extrinsic impurities as a major contributor.","The two codes disagree on tungsten migration: kinetic tracing moves W toward the X-point through the $\\nabla B$ drift, while the no-drift fluid version needs a 100-fold lower anomalous diffusivity to produce similar profiles."],"supporting_citations":[{"why":"Supplies the fluid edge-plasma solver used to produce the validated helium background and the multi-species simulations including tungsten.","marker":"[9, 10]"},{"why":"Supplies the kinetic Monte-Carlo impurity transport code used as the reference for erosion and tungsten migration.","marker":"[11]"},{"why":"Provides the sheath potential formula used to compute ion impact energies in the analytical model and in ERO2.0.","marker":"[14]"},{"why":"Provides the He-on-W sputtering yields used in the analytical erosion estimate.","marker":"[15]"},{"why":"Provides the SDTrimSP sputtering simulations used for the He-W yield database in ERO2.0.","marker":"[16]"},{"why":"Supplies the post-mortem measurements of exposed tungsten samples that set the experimental erosion benchmark.","marker":"[22, 23]"},{"why":"Provides the 'jett' tungsten bundling model used to include 24 bundled W species in SOLPS-ITER.","marker":"[31]"},{"why":"Provides the He-W and W-W sputtering data used in the SOLPS-ITER tungsten source calculation.","marker":"[33]"}],"fun_headline_variants":["He2+ ions, not He+, drive tungsten erosion in tokamak divertors","Fluid boundary condition inflates tungsten erosion 20x in simulations","Kinetic modelling corrects tungsten source overestimate in helium plasmas","Self-sputtering explains 20-fold tungsten erosion overestimate in tokamaks","He2+ dominates tungsten sputtering even at 50-50 flux mix"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything hinges on the unvalidated SOLPS-ITER ion temperature and on the constant extrapolation of plasma parameters from the simulation grid to wall surfaces; if those near-wall values are wrong, the predicted erosion rates and the inferred role of impurities change substantially.","fun_headline_variants_meta":{"raw":{"variants":["He2+ ions, not He+, drive tungsten erosion in tokamak divertors","Fluid boundary condition inflates tungsten erosion 20x in simulations","Kinetic modelling corrects tungsten source overestimate in helium plasmas","Self-sputtering explains 20-fold tungsten erosion overestimate in tokamaks","He2+ dominates tungsten sputtering even at 50-50 flux mix"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00029,"raw_usage":{"total_tokens":1751,"prompt_tokens":1056,"completion_tokens":695,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":672,"completion_tokens_details":{"reasoning_tokens":595}},"tokens_in":672,"tokens_out":695,"duration_ms":7195,"temperature":1.0,"reasoning_tokens":595,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T10:52:47.362566+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a repeat discharge with the same settings as #36687, measure the divertor ion temperature and O6+ density simultaneously, and compare ERO2.0's predicted erosion thickness against post-mortem marker measurements: if measured O6+ is below about 1% and erosion is still 50-100 nm, the model chain is missing something; if O6+ is a few percent and erosion matches, the impurity explanation is confirmed.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the kinetic Monte-Carlo impurity transport code used as the reference for erosion and tungsten migration."},{"cited_title":"Philadelphia, Pennsylvania)","cited_arxiv_id":null,"evidence_quote":"Provides the sheath potential formula used to compute ion impact energies in the analytical model and in ERO2.0."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the He-on-W sputtering yields used in the analytical erosion estimate."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the SDTrimSP sputtering simulations used for the He-W yield database in ERO2.0."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the 'jett' tungsten bundling model used to include 24 bundled W species in SOLPS-ITER."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the He-W and W-W sputtering data used in the SOLPS-ITER tungsten source calculation."}],"review_version":1}