{"id":"030e178c-48fa-46a3-91a7-c8021f9f24c2","arxiv_id":"2411.14640","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A new equilibrium chemistry solver for inner protoplanetary discs includes thermionic and ionic emission from grains with arbitrary size distributions, and shows grain charging can freeze the grain size distribution.","lead":"This paper builds a chemical model of the hot inner regions of planet-forming discs that combines dust grains of all sizes with the emission of electrons and ions from grain surfaces. It shows that the standard shortcut for treating grain sizes gives wrong grain charges, which changes predictions for how fast dust grains collide and grow.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Freeze-in conclusion rests on setting K_ij=0 below the Coulomb barrier; the acknowledged Brownian tail is never quantified in Figs. 13-14.","rationale":"The reader identified the same load-bearing weakness: the collision kernel truncation in Section 2.4.2 is the most fragile premise under the paper's strongest quantitative conclusion. I agree with that choice. The paper is a credible, well-validated extension of earlier work: Section 3.1.1 reproduces Desch & Turner (2015) in the single-size limit, Section 3.1.2 matches Marchand et al. (2022) over the full temperature range, and Appendix C demonstrates that the single-charge-per-size approximation is consistent with a Gaussian charge distribution. The effective-dust-to-gas-ratio failure and the general qualitative effect of charging on collision time-scales are supported by the presented calculations. The concern is not that the model is wrong, but that the freeze-in conclusion is stated without a sensitivity test for the acknowledged Brownian tail and the mean-charge evaluation of an exponentially nonlinear kernel. Because the paper itself flags the K_ij = 0 approximation, the appropriate outcome is to require a concrete test of that approximation before the freeze-in statement is used as a firm prediction. This does not move the reader's conditional verdict; it sharpens the condition under which the conclusion should be accepted.","tokens_in":33010,"tokens_out":4488,"duration_ms":51418,"concrete_test":"Recompute the charged-grain collision time-scale in Fig. 13 for representative pairs (e.g., a_i = a_j = 10^-5, 10^-4, 10^-3 cm) at T = 800, 1000, and 1200 K and n_H2 = 10^14 cm^-3, using the full Okuzumi et al. (2011a) kernel integral over the Maxwellian relative-velocity distribution around the drift velocity instead of setting K_ij = 0 below the barrier. If feasible, also Monte-Carlo sample the per-size charge distributions from Appendix C rather than using only the mean charge. Compare the resulting shortest collision time-scale with the local dynamical time t_dyn = 1/Omega_K. If the ratio remains above ~10^2 for all unshaded temperatures, the freeze-in claim is robust; if it drops below ~10 for any representative case, Section 3.6's freeze-in statement must be softened to depend on the zero-kernel and mean-charge approximations.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's headline claim that grain charging 'significantly alters grain collisional time-scales' and, in Section 3.6, that the grain size distribution may be 'frozen in' at the inner-disc mid-plane is computed from Eq. (33) using the limiting kernels (35a)-(35b). In Section 2.4.2 the authors state that, following Akimkin et al. (2023), they set K_ij = 0 whenever KE_D_ij < U(a_i + a_j), with the explicit acknowledgment that a nonzero Brownian contribution always persists. The freeze-in conclusion is therefore exactly as strong as the uncomputed tail of the Maxwellian relative-velocity distribution. The text says the plotted 10^40 s plateau is a truncation and the real time-scales are 'even higher'; that statement relies on the omitted tail being utterly negligible. This is plausible when U/kT >> 1, but for small grains near the charging transition, or for pairs with drift energies just below the barrier, the exponential suppression is not necessarily severe enough to keep the collision time-scale orders of magnitude above the dynamical time. A second, related sensitivity is that the kernel is evaluated at a single charge per grain size; Appendix C shows the charge distribution is Gaussian, but the collision kernel depends exponentially on Z_i Z_j, so the mean-charge calculation could in principle miss a contribution from the distribution tails. Neither issue is an internal contradiction, but both sit directly under the strongest quantitative conclusion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript presents an equilibrium chemical network for the inner protoplanetary disc that combines non-thermal and thermal ionization, gas-phase recombination and charge transfer, and grain charging with ionic and thermionic emission for an arbitrary grain size distribution. The authors solve the nonlinear system with a successive-over-relaxation plus Powell hybrid method, validate against Desch & Turner (2015) in the single-size limit and against Marchand et al. (2022) with emission switched off, and use the resulting abundances to compute resistivities and grain-grain collision time-scales. They report that the effective dust-to-gas ratio approximation yields inaccurate grain charges, that grain charge is proportional to size across distributions, and that electrostatic repulsion can make grain collision time-scales far exceed the dynamical time, potentially freezing the size distribution in the inner disc.","tokens_in":33265,"tokens_out":6317,"duration_ms":66070,"significance":"If the results hold, the paper fills a real gap: it is the first to combine arbitrary grain size distributions with ionic and thermionic emission in an equilibrium inner-disc network, and it provides a widely usable numerical method. The validation against two prior codes is a substantial strength, as is the analytic Appendix C derivation of the Gaussian charge distribution and the Z-proportional-to-a scaling. The effective-dust-to-gas-ratio comparison and the resistivity calculations give the community practical tools. However, the quantitative freeze-in conclusion is currently tied to a deliberately truncated collision kernel and to mean-charge evaluation; those points must be fortified before the headline claim can be fully trusted.","major_comments":[{"comment":"The freeze-in conclusion is computed with K_ij set to zero whenever KE_D_ij < U(a_i+a_j), even though the text acknowledges that a nonzero Brownian contribution always persists. This makes the reported collision time-scales upper limits, and the statement in Section 3.6 that the size distribution 'may be frozen in' is exactly as strong as the unquantified tail of the relative-velocity distribution. Please provide a quantitative estimate or a bounding argument for the omitted contribution, particularly for pairs whose drift energy is only slightly below the Coulomb barrier and for temperatures near the charge-sign transition. A related approximation is that Eq. (35b), formally valid only for KE_D_ij much larger than U(a_i+a_j), is used over the entire range KE_D_ij > U; the sign and rough magnitude of the error this introduces should also be stated. If the neglected tail contributes non-negligible collision rates, the magnitude of the claimed effect must be revised.","section":"Section 2.4.2, Eqs. (35a)-(35b); Figs 13-14"},{"comment":"The single-charge-per-grain-size approximation is validated only for the mean charge and the Gaussian width. Since the collision kernel is exponentially sensitive to Z_i Z_j, evaluating the kernel at the mean charge is not equivalent to averaging over the charge distribution. Please add a sensitivity estimate, using the Appendix C distributions, of the collision rate averaged over the charge distribution, e.g., <exp(-Z_i Z_j e^2/[(a_i+a_j) kT])>, versus the same quantity evaluated at the mean charges for representative pairs. This is directly load-bearing for the claimed order-of-magnitude changes in collision time-scales.","section":"Section 2.3 and Appendix C"},{"comment":"The claim that the effective dust-to-gas ratio method yields 'significantly inaccurate grain charges' is supported by a visual comparison of nine panels, but no fractional error is quantified. Since this is one of the two headline claims in the abstract, please add a quantitative error measure, such as the maximum or rms fractional error in Z_i as a function of temperature and grain size, for the cases shown in Fig. D1.","section":"Appendix D, Fig. D1"}],"minor_comments":[{"comment":"The normalization integral appears to have the same upper limit a_max in both terms of the denominator; presumably the second term should involve a_min. Please correct this typo.","section":"Eq. (7)"},{"comment":"The number density is written as 'nH2 = 1014gcm^-3' and should be cm^-3; likewise the collision cross-section in Eq. (37) is written as cm^-2 and should be cm^2.","section":"Section 2.2 and Eq. (37)"},{"comment":"The comparison with Desch & Turner (2015) uses a non-thermal rate zeta = 1.4e-22 s^-1, whereas the fiducial runs use 7.6e-19 s^-1; the text explains the historical origin but a sentence clarifying why the comparison value is not used elsewhere would help.","section":"Section 3.1.1"},{"comment":"The captions mention a plotted plateau of 10^40 s, but this value is not apparent from the axis ranges shown in the main text; consider an inset or an explicit annotation so the truncation is clear.","section":"Figs 13 and 14"}],"recommendation":"major_revision","confidential_remarks":"The paper is within the scope of the journal and makes a useful methodological contribution. The main technical concern is the collision-kernel truncation and the mean-charge evaluation, both of which sit directly under the freeze-in claim; these are fixable with additional sensitivity calculations and do not require recasting the network itself."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper is a legitimate advance for inner-disc ionization chemistry. It is the first equilibrium network I know that includes both ionic/thermionic emission from grains and an arbitrary grain size distribution, and it validates cleanly against Desch & Turner (2015) in the single-size limit and Marchand et al. (2022) when emission is switched off. The demonstration in Appendix D that the effective dust-to-gas-ratio method gives wrong grain charges even when gas-phase abundances are fine is well argued and will matter for anyone doing coagulation/fragmentation work. Appendix C, where they derive the Z ∝ a scaling with emission from the master equation, is a genuinely useful analytic result.\n\nThe soft spots are real but not fatal. The headline claim about frozen grain size distributions in Section 3.6 rests on the collision kernel being set to zero when the drift kinetic energy is below the Coulomb barrier, following Akimkin et al. (2023). The authors acknowledge that a Brownian contribution always exists, but they never quantify it. For pairs with drift energies just below the barrier, or small grains near the charging transition, the exponential suppression may not be steep enough to keep collision timescales orders of magnitude above the dynamical time. The plotted 10^40 s plateau is a truncation; the true timescales are “even higher” only if the omitted tail is negligible. That needs a sensitivity test, not just an assertion.\n\nThe second soft spot is the single-charge-per-grain-size approximation. Appendix C shows the charge distribution is Gaussian, and the mean-charge calculation is likely fine for abundances, but the collision kernel depends exponentially on Z_i Z_j. Tails of the distribution could in principle contribute more to collision rates than the mean suggests. Again, addressable with a simple test.\n\nMinor issues: no code or data are shipped (though the method is standard enough to reproduce), and there is a small inconsistency in how the shaded regions in Figs. 13–14 are described in the text.\n\nThis paper deserves a serious referee. It fills a genuine gap, the validation is careful, and the conclusions are potentially important for disc evolution. The main quantitative claim needs reinforcement, not rethinking. I would send it to peer review with a request for sensitivity runs on the collision kernel and a clear statement of where the freeze-in conclusion holds.","headline":"First combined treatment of ionic/thermionic emission and arbitrary grain-size distributions; solid validation, but the freeze-in conclusion needs sensitivity tests on the collision kernel.","tokens_in":33824,"tokens_out":2346,"would_cite":true,"duration_ms":22539,"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":"Grain charging, computed across all dust sizes with ionic and thermionic emission, can freeze the inner disc's grain population.","keywords":["protoplanetary discs","ionization chemistry","grain charging","thermionic emission","ionic emission","grain size distribution","grain collisions","resistivities"],"falsifier":"Evaluate the full collision integral of Eq. (34)—including the Maxwellian tail at $\\Delta v_{ij}^{D} < \\Delta v_{ij}^{Br}$—with the same grain charges and disc conditions as in Fig. 13, and ask whether the shortest grain–grain collision time-scale ever drops below the dynamical time in the unshaded 800–1500 K region; a drop would falsify the freeze-in conclusion. A complementary check is to recompute collision time-scales using a distribution of charges per grain size (rather than a single charge) and a dipole-inclusive cross-section, since both are omitted in the paper.","tokens_in":32751,"feed_emoji":"🪐","tokens_out":6421,"duration_ms":58069,"temperature":0.7,"pith_summary":"The paper argues that in the inner regions of protoplanetary discs, grain charging cannot be treated with a single grain size or a rescaled dust-to-gas ratio. It presents an equilibrium chemical network that simultaneously includes gas-phase ionization, ionic and thermionic emission from grain surfaces, and an arbitrary distribution of grain sizes, each bin carrying its own charge. With this network the authors show that the commonly used 'effective dust-to-gas ratio' method, while roughly reproducing gas-phase charge densities, predicts grain charges that can be wrong by an order of magnitude or more. They further show that the resulting electrostatic repulsion makes grain–grain collision time-scales vastly longer than the local dynamical time at the 1 au mid-plane, so a grain size distribution entering the inner disc may be effectively frozen in. If correct, this changes how disc viscosity, dust coagulation and fragmentation, and opacity should be modelled in the inner disc.","feed_headline":"Charged dust may freeze the inner disc's grain sizes","feed_subtitle":"A network that includes every grain size fixes charges the old shortcut gets wrong.","key_machinery":"The central object is the equilibrium reaction network of equations (14*)–(17*), a nonlinear system in the per-size grain charges $Z_i$, electron, molecular ion and alkali ion densities, closed by charge conservation. The load-bearing simplification is the 'reduced temperature' $\\tau \\equiv akT/e^2$: because the Draine–Sutin focusing factors and the Saha–Langmuir ion-emission fraction $f_+$ depend on charge only through $Z/\\tau$, the most probable charge on every size bin is $Z_i = \\psi \\tau_i$ with a single constant $\\psi$ set by one transcendental equation (Appendix C). This reduces an $N$-dimensional problem to a root-finding problem and is why the paper can handle arbitrary size distributions. The analysis of grain collisions relies on the limiting collision kernels of Okuzumi et al. (2011a) for Coulomb-interacting grains, used in the truncated form of Akimkin et al. (2023): the kernel is set to zero when the drift kinetic energy falls below the Coulomb barrier $U(a_i+a_j)$, which is precisely the assumption that produces the long freeze-in collision time-scales.","core_discovery":"The paper's central claim is that a chemical network that includes ionic and thermionic emission from grains and a full, arbitrary grain size distribution—solved exactly in equilibrium—yields results that differ materially from prior treatments. Specifically, the paper shows that approximating the grain population by a single size with an 'effective dust-to-gas ratio,' as in earlier work, reproduces the dominant gas-phase ionization fractions (and hence resistivities) but gives significantly erroneous grain charges, particularly for flatter size distributions. It further shows that grain charging, which becomes severe once grain-surface ion emission and thermionic emission set in above roughly 500–600 K, lengthens grain–grain collision time-scales by many orders of magnitude relative to the dynamical time at the disc mid-plane near 1 au, implying that the grain size distribution can be frozen in while grains traverse the inner disc. Along the way the authors establish that the mean grain charge remains proportional to grain size, $Z \\propto a$ (equivalently a constant $\\psi \\equiv Z/\\tau$ with reduced temperature $\\tau = akT/e^2$), even when ionic and thermionic emission are active, so that a single root-finding exercise fixes the charge on every size bin.","pith_inferences":["If the freeze-in result survives a full collision kernel, large grains that form outside the inner disc could pass through the 1 au region without being eroded by fragmentation, changing the opacity and the location of the dead-zone edge; this is an inference, since the paper does not evolve the size distribution.","The single-charge-per-size simplification is justified by near-Gaussian charge distributions, but the collision kernel truncation and the neglect of dipole interactions are exactly the places where a distribution of charges could reopen collisions; a direct Monte Carlo or full-kernel test would bound the freeze-in effect.","The $\\psi$-scaling implies a charge-to-mass ratio that grows with grain size, which could affect settling and drift signatures: small grains carry little charge per mass, large grains much more; this is not explored in the paper.","The method's structure—one root equation fixing all grain charges—should extend naturally to multi-alkali networks (the paper demonstrates sodium alongside potassium) and to any disc model that supplies $T$, $n_{\\rm H2}$ and a grain size distribution, so it is a candidate drop-in module for coagulation–chemistry co-evolution codes; that application is left implicit."],"forward_implications":["The effective dust-to-gas-ratio method with $p=1.5$ remains adequate for gas-phase charge densities and thus for resistivities at $T \\gtrsim 1000$ K, but it is not safe for any calculation that depends on the charge of individual grains.","At the 1 au mid-plane, charged-grain collision time-scales exceed the dynamical time by many orders of magnitude over most of the 800–1500 K range, so an inner-disc grain size distribution can be frozen in unless it collided on the way in.","Grain charge is proportional to grain size, $Z \\propto a$, for all temperatures considered, extending the Draine–Sutin linear law to the regime where ionic and thermionic emission are active.","Chemical equilibrium is a good approximation in the inner disc: the chemical time-scale is shorter than the dynamical, thermal and grain-collision time-scales in the region of interest, with the no-potassium network setting the relevant time-scale below ~900 K.","Ohmic, Hall and ambipolar resistivities can now be computed from a network that includes emission and a full size distribution; the sign of the Hall term flips when grains replace electrons as the dominant negative charge carriers."],"supporting_citations":[{"why":"Provides the single-grain-size network with ionic and thermionic emission that this work generalizes to a full grain size distribution, including the potassium condensation and emission parameters.","marker":"Desch & Turner (2015)"},{"why":"Introduced the effective dust-to-gas-ratio approximation with p=1.5 that the paper tests and finds inaccurate for grain charges.","marker":"Jankovic et al. (2021)"},{"why":"Supplies the equilibrium network with a distribution of grain sizes but no emission, used as the validation baseline for the new network's abundances.","marker":"Marchand et al. (2022)"},{"why":"Gives the electrostatic focusing factors and the linear grain-charge versus size law for reduced temperature above unity that the paper extends to emission-active conditions.","marker":"Draine & Sutin (1987)"},{"why":"Derives the limiting forms of the collision kernel for Coulomb-interacting grains that the paper uses to compute grain–grain collision time-scales.","marker":"Okuzumi et al. (2011a)"},{"why":"Provides the coagulation model with the truncated kernel (collision rate set to zero below the Coulomb barrier) that the paper adopts and compares against for inner-disc grain survival.","marker":"Akimkin et al. (2023)"},{"why":"Established the effective dust-to-gas-ratio concept for charged grains in cooler disc regions, the approach Appendix D shows fails for grain charges when emission is active.","marker":"Bai & Goodman (2009)"},{"why":"Supplies the method used to compute Ohmic, Hall and ambipolar resistivities from the network's ionization fractions.","marker":"Wardle (2007)"}],"fun_headline_variants":["Old shortcut gets grain charges wrong; new network fixes it","Full grain size network reveals charging freezes inner disc dust","Ionic and thermionic emission alter grain charge, freeze sizes","Grain size distribution key to charging, may lock dust sizes","New exact network: grain charging can stall dust growth in disc"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The collision-rate results rest on setting the grain–grain collision kernel to zero whenever the drift kinetic energy is below the Coulomb repulsion barrier, even though a finite contribution from Brownian (Maxwellian) relative motion always remains; if that residual kernel keeps collisions frequent, the claimed freeze-in of the size distribution would be weaker than stated.","fun_headline_variants_meta":{"raw":{"variants":["Old shortcut gets grain charges wrong; new network fixes it","Full grain size network reveals charging freezes inner disc dust","Ionic and thermionic emission alter grain charge, freeze sizes","Grain size distribution key to charging, may lock dust sizes","New exact network: grain charging can stall dust growth in disc"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000349,"raw_usage":{"total_tokens":1973,"prompt_tokens":1080,"completion_tokens":893,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":696,"completion_tokens_details":{"reasoning_tokens":809}},"tokens_in":696,"tokens_out":893,"duration_ms":8584,"temperature":1.0,"reasoning_tokens":809,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:04:06.088221+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the full collision integral of Eq. (34)—including the Maxwellian tail at $\\Delta v_{ij}^{D} < \\Delta v_{ij}^{Br}$—with the same grain charges and disc conditions as in Fig. 13, and ask whether the shortest grain–grain collision time-scale ever drops below the dynamical time in the unshaded 800–1500 K region; a drop would falsify the freeze-in conclusion. A complementary check is to recompute collision time-scales using a distribution of charges per grain size (rather than a single charge) and a dipole-inclusive cross-section, since both are omitted in the paper.","supporting_citations":[{"cited_title":"V., Caselli P., Gong M., Silsbee K., 2023, @doi [ ] 10.3847/1538-4357/ace2c5 , https://ui.adsabs.harvard.edu/abs/2023ApJ...953...72A 953, 72","cited_arxiv_id":null,"evidence_quote":"Provides the coagulation model with the truncated kernel (collision rate set to zero below the Coulomb barrier) that the paper adopts and compares against for inner-disc grain survival."}],"review_version":1}