{"id":"8e0f6849-2d42-4108-9557-02d4b6af00e7","arxiv_id":"2608.03998","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Two electrons can escape an atom simultaneously through a collective tunneling channel, with the electron-electron repulsion shaping the escape path and leaving a measurable lateral momentum signature.","lead":"This paper claims that in a strong laser field two electrons can tunnel out of an atom together, rather than one after the other. It derives a theory for this 'collective tunneling' and predicts a sideways split in the electrons' momentum that experiments could look for.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Analytic rate formula (48) overestimates the action by a factor of several at the fields used in the TDSE, so the central quantitative claim is not supported in the demonstrated regime.","rationale":"The reader's conditional acceptance is appropriate. My stress-test sharpens the reader's weakest_assumption by showing that the nuclear attraction is not a small correction in the regime where the theory is compared to TDSE. The paper's own numbers (ImS_min^(1)=1.18 vs Eq. (51) approximately 4.7) imply the analytic rate formula overestimates the action by a factor of four in the exponent at E=0.035, corresponding to roughly three orders of magnitude in the rate. This does not disprove the existence of collective tunneling—the TDSE still shows a diagonal flux and lateral anti-correlation—but it means the paper's central quantitative claim (Eq. 48) is not established at the demonstrated parameters. Since the paper explicitly acknowledges the semiclassical expansion is at the edge of its validity and lacks a prefactor, conditional acceptance remains the correct verdict. No change to the reader's verdict is needed.","tokens_in":20949,"tokens_out":28866,"duration_ms":280458,"concrete_test":"Use the data already in Fig.5: at E=0.035, ImS_min^(1)=1.18 for Br^-. Evaluate Eq. (51) for the same Br^- parameters (I_p1=3.37 eV, I_p2=11.81 eV, Z=1). If the analytic ImS exceeds the numerical value by more than a factor of 2, then the Coulomb correction is not a small perturbation and Eq. (48) cannot be used for quantitative rate predictions at E/E_BS=0.83. Repeat at E=0.02 where ImS_min^(1)=7.24; if the ratio stays above 1.3, the analytic formula is asymptotically valid only for E << E_BS.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing weakness is the perturbative treatment of the electron-nuclear attraction in the analytic rate. Section V drops the nuclear term from the sub-barrier potential (Eq. 28) and later adds it as a first-order correction S_C (Eqs. 44–47). This requires |S_C| << S_0 and sqrt(mu)*Lambda_CT << 1, which fails precisely at the parameters used for the TDSE and the experimental predictions. For Br^- (I_p=15.18 eV=0.558 a.u., Z=1), the paper's own numerical minimization gives ImS_min^(1)=1.18 at E=0.035 (E/E_BS=0.83) and 7.24 at E=0.02. Evaluating Eq. (51) for the same parameters gives ImS approximately 4.7 and 10.6, respectively—a factor of 4 and 1.5 in the exponent, corresponding to many orders of magnitude in the rate. The discrepancy arises because the nuclear attraction creates a classically allowed region near the nucleus, drastically shortening the sub-barrier path; this is a non-perturbative effect, not a small correction. The paper acknowledges the semiclassical expansion is 'on the edge of applicability' but still presents Eq. (48) and Eq. (59) as quantitative predictions for the regime E/E_BS~0.8. Thus the central rate formula is not quantitatively supported where the collective channel is demonstrated; only the qualitative picture survives.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a combined TDSE and semiclassical study of simultaneous two-electron tunnel ionization (collective tunneling) in strong low-frequency laser fields. Two-dimensional TDSE simulations for Br^- and Xe in short unipolar pulses show a two-electron flux along the diagonal x1≈x2 with lateral anti-correlation y1≈-y2, which is absent in 1D and disappears at longer pulse durations. The authors interpret this as tunneling through a two-dimensional barrier and develop an imaginary-time semiclassical theory for the symmetric trajectory. The main analytic results are the collective-tunneling action S_CT = S_0 + S_y + S_C (Eq. 48), with S_y arising from electron-electron repulsion and S_C from the nuclear attraction, and the asymptotic lateral momentum prediction p_y ≈ (π^2 E^2/(16 I_p))^{1/6} (Eq. 59). The paper further proposes experimental signatures of collective tunneling in circularly polarized or short pulses.","tokens_in":21327,"tokens_out":8013,"duration_ms":95816,"significance":"If the analytic theory were quantitatively reliable, this would be an important step: it gives a parameter-free semiclassical description of a correlated two-electron tunneling channel, corrects the earlier quasiparticle models by including electron-electron repulsion, and produces a falsifiable prediction for lateral momentum distributions. The TDSE evidence for the collective channel and for the lateral shoulder is direct and not fitted to the analytic model; the lateral-momentum comparison is a genuine consistency check. The limitations of the semiclassical expansion are acknowledged in the text, but the quantitative rate formula is used for experimental estimates, so the status of the analytic predictions is a key issue.","major_comments":[{"comment":"The quantitative rate formula is not supported at the field strengths used in the TDSE demonstration. For Br^- at E=0.035, the paper's own numerical minimization of the full action gives ImS_min^(1)=1.18 (and ImS_min^(2)=1.228 with the unsmoothed potential), while Eq. (51) gives ImS_CT ≈ 4.7. At E=0.02 the corresponding numbers are 7.24 and ≈10.6. Thus the analytic exponent is larger by factors of about 4 and 1.5, corresponding to many orders of magnitude in rate. This contradicts the statement in Sec. V (note after Eq. 49) that the perturbative results \"demonstrate a good agreement with those obtained without perturbative expansions as well as with the TDSE solutions.\" Since Eq. (48) is used in Sec. VI.A to compare CT and sequential rates and to estimate experimental feasibility, the central quantitative claim is not established in the demonstrated regime.","section":"Sec. V, Eq. (51) and Fig. 7"},{"comment":"The derivation rests on a perturbative treatment of the nuclear attraction that is not controlled at the parameters of interest. The term -2Z/sqrt(x^2+y^2) is dropped in Eq. (28) using y0/x0~sqrt(mu)<<1 (Eq. 45), but for Br^- at E=0.02 the paper's own Eq. (43) gives mu=0.61, so sqrt(mu)~0.8, not <<1. The resulting short-range action (42) at E=0.035 is about 15, whereas the full numerical action is 1.18; the first-order Coulomb correction (47) reduces this only to about 4.7. The near-nucleus region, where |y|>=|x| and the integrand in Eq. (46) should be 1/|y| rather than 1/|x|, contributes nonperturbatively, as the paper partly acknowledges in comment 2 of Sec. V.B. This is not a small correction and requires either a nonperturbative treatment or a strictly asymptotic statement of validity.","section":"Sec. V.A and V.B, Eqs. (28), (42), (47)"},{"comment":"The lateral momentum prediction p_y=(π^2 E^2/(16 I_p))^{1/6} is derived from the asymptotic exit coordinate y0 of Eq. (37), i.e. from the short-range approximation without the nuclear attraction. The TDSE position of the side maxima agrees only to about 20%, and the improvement is obtained by solving the full equations of motion (58) with a numerical initial condition. The paper should state more clearly that Eq. (59) is a leading-order estimate, not a quantitative prediction at E/E_BS~0.8, and should quantify the expected error or give a parameter scan before using it as an experimental calibration.","section":"Sec. VI.B, Eq. (59)"}],"minor_comments":[{"comment":"The text refers to \"Fig.2(c.1–c.3)\" for the xenon panels, but the figure uses panels (g)-(i).","section":"Sec. IV"},{"comment":"Typo: \"anlyze\" should be \"analyze\".","section":"Sec. III"},{"comment":"Typo: \"the very fact the the action\" should read \"the very fact that the action\".","section":"Sec. V"},{"comment":"\"Not that in the 3D geometry\" should be \"Note that in the 3D geometry.\"","section":"Sec. VI.C"},{"comment":"The text around Fig. 11 states the field propagates along the y-axis, while the rest of the paper uses x as the polarization direction; please clarify the coordinate convention in the figure caption.","section":"Sec. VI.C, Fig. 11"}],"recommendation":"major_revision","confidential_remarks":"The TDSE part is convincing and should be preserved. The main weakness is that the analytic rate formula, despite caveats, is used for quantitative experimental estimates in a regime where it differs from the authors' own numerical minimization by several units of action, i.e. orders of magnitude in rate. The paper would need either to restrict Eqs. (48)/(51) and Fig. 7 to the asymptotic weak-field regime, or to replace the perturbative Coulomb treatment with a nonperturbative evaluation of the nuclear contribution. The lateral momentum prediction is on stronger footing but should be labeled an estimate. I would support publication after this is addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is the S_y term: the analytic account of e-e repulsion in the sub-barrier action, which corrects the old Zon/Eichmann quasiparticle picture by orders of magnitude. That correction is parameter-free and is not fitted to the TDSE. The 2D TDSE shows a diagonal flux and lateral momentum shoulders in both Br- and Xe, which is direct evidence that something collective is happening. Credit where due: the paper identifies the right physical mechanism (lateral separation y1 = -y2 to minimize the barrier) and gives the field a measurable signature (p_y ~ (pi^2 E^2/16 I_p)^{1/6}). The comparison to the non-interacting-electron TDSE control is the right way to isolate the e-e effect.\n\nThe soft spot is the one the stress-test note flags, and it is real. The analytic rate formula (48) is built on dropping the nuclear term in the sub-barrier potential and then adding it back perturbatively. That requires both |S_C| << S_0 and sqrt(mu)*Lambda_CT << 1. At E = 0.035 for Br-, the paper's own numeric minimization gives Im S_min^(1) = 1.18, while Eq. (51) gives about 4.7. A factor of 4 in the exponent is many orders of magnitude in the rate. The nuclear attraction is not a small correction at E/E_BS ~ 0.8; it creates a classically allowed region near the nucleus that shortens the path. The paper states the expansion is 'on the edge of applicability' but still presents (48) and (59) as quantitative. That is the load-bearing flaw: the exponential rate is not supported in the demonstrated regime. The qualitative picture—collective channel exists, is 2D, has a lateral momentum shoulder—survives. The momentum prediction (59) is less sensitive to the Coulomb correction than the rate, and the TDSE shoulder position agrees to ~20%, so that part holds up better.\n\nAlso minor: no prefactor for the rate (acknowledged), 2D TDSE without convergence tests in grid size or absorber, and the experimental proposal is schematic. None of these are fatal on their own.\n\nWho is this for? Strong-field atomic physics people who care about NSDI mechanisms and tunneling theory. It deserves a serious referee because it identifies a new channel and fixes an old quantitative error, even though the central rate formula needs a non-perturbative treatment of the nuclear attraction before I would trust its numbers. I would not cite the rate formula as it stands, but I would cite the qualitative result and the momentum signature.","headline":"A serious attempt at a real two-electron tunneling theory with a new repulsion-induced action term and a plausible momentum signature, but the analytic rate is not quantitatively controlled at the fields where the TDSE evidence lives.","tokens_in":21791,"tokens_out":665,"would_cite":true,"duration_ms":8799,"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":"Two electrons can tunnel out of an atom together, and this paper derives the rate and the observable signature.","keywords":["collective tunneling","nonsequential double ionization","tunnel ionization","electron–electron correlations","strong-field laser physics","imaginary time method","lateral momentum distribution","negative bromine ion"],"falsifier":"Measure two-electron coincidence lateral momenta for nonsequential double ionization of xenon in a short circularly polarized pulse with recollisions suppressed. If the collective channel exists, the p1y = −p2y diagonal should show side peaks or shoulders near p_y = (π^2 E^2/(16 Ip))^(1/6); a distribution identical to the non-interacting sequential background at the same field would refute the prediction. Additionally, at fields E much below the barrier-suppression field, a direct numerical evaluation of Im S along the full two-particle trajectory should reproduce Eq. (48); a mismatch there wo","tokens_in":20882,"feed_emoji":"⚛️","tokens_out":5788,"duration_ms":67431,"temperature":0.7,"pith_summary":"The paper argues that in strong low-frequency laser fields two electrons can tunnel out of an atom simultaneously, not one after the other, and that this collective channel leaves a measurable fingerprint. It develops an analytic semiclassical theory that includes electron–electron repulsion, which earlier collective-tunneling models neglected, and checks it against numerical solutions of the two-electron Schrödinger equation for negative bromine and neutral xenon. The most probable escape path keeps the two electrons at the same coordinate along the field while separating them in opposite lateral directions; the repulsion then imprints a characteristic lateral momentum on the pair. If the theory is right, collective tunneling becomes a quantitative, experimentally searchable process rather than a speculative correction to sequential ionization.","feed_headline":"Two electrons can tunnel out together; theory gives the rate","feed_subtitle":"Collective tunneling leaves side peaks in electron-pair momenta, an observable fingerprint for short laser pulses.","key_machinery":"The central object is the symmetric two-electron trajectory ansatz x1 = x2 ≡ x, y1 = −y2 ≡ y. On this slice the two-electron problem becomes a single effective particle with mass 2 in the potential V(x,y) = 2V_ei(x,y) + V_ee(x,y;x,−y) − 2E x. The semiclassical imaginary-time action separates into a longitudinal part Sx and a lateral part Sy; the lateral equation of motion y¨ = −1/(4y^2) encodes the electron–electron repulsion and yields both the sub-barrier action correction and, after real-time propagation, the asymptotic lateral momentum. The Coulomb attraction to the nucleus is then added perturbatively as SC through the standard matching procedure. The work this machinery does is to conv","core_discovery":"For strong fields below the barrier-suppression limit, the paper claims that nonsequential double ionization contains a collective tunneling channel in which both electrons pass under the barrier at once. Along the dominant trajectory the electron positions satisfy x1 ≈ x2 and y1 ≈ −y2, reducing the two-electron problem to a single particle of mass 2 moving in an effective potential. The electron–electron repulsion, kept as 1/(2|y|), dominates the nuclear attraction in the leading sub-barrier action and contributes the term Sy ≈ i 3π^(2/3)/(2^(4/3)) (Ip/E^2)^(1/6); adding the Coulomb factor gives the total action S_CT = S_0 + S_y + S_C and the rate w_CT ≃ exp(−2 Im S_CT). After the electrons","pith_inferences":["The scaling p_y ∝ E^(1/3) Ip^(−1/6) is a clean cross-species test: measuring the side-peak position at several field amplitudes and atomic targets would directly check the symmetric-path assumption on which the whole derivation rests.","In a real three-dimensional atom, CT should emit the two electrons along a cone around the field axis rather than exactly in a plane; a coincidence experiment seeking pairs with nearly equal energy and opposite polar angles could detect CT even at low yield.","Because the theory splits the action into short-range (electron–electron repulsion) and Coulomb parts, the same decomposition may transfer to other correlated tunneling problems in which a repulsive interaction dominates along one coordinate.","The authors note that 1D model atoms show no collective channel at all; this implies that any simulation or experiment searching for CT must preserve at least two-dimensional geometry for the electron motion."],"forward_implications":["The exponential rate of collective tunneling is specified by Eq. (48), so experiments and simulations can compare CT and sequential rates directly from ionization potentials, field amplitude, and atomic charge.","The lateral momentum prediction p_y ≈ (π^2 E^2/(16 Ip))^(1/6) gives an unambiguous signature: side peaks for Br− and shoulders for Xe in the p1y = −p2y diagonal of the two-electron momentum distribution.","Short circularly polarized or unipolar pulses that suppress recollision can expose CT, because the pair emerges with correlated off-plane momenta rather than in the polarization plane.","Including electron–electron repulsion removes the orders-of-magnitude overestimate of earlier quasiparticle CT models, bringing the CT rate to a level comparable to the sequential channel under short-pulse conditions.","The same imaginary-time action decomposition extends the standard single-electron tunnel-ionization formalism to a correlated two-particle problem, with potential applications to other two-electron escape processes."],"supporting_citations":[{"why":"Original quasiparticle picture of collective tunneling whose rate the paper corrects by including electron–electron repulsion.","marker":"[19]"},{"why":"Early collective-tunneling model that the paper shows is inconsistent with numerical calculations when repulsion is discarded.","marker":"[20]"},{"why":"Previous numerical TDSE study that identified the collective flux in short pulses and supplies the non-interacting-electron comparison distributions.","marker":"[32]"},{"why":"Perelomov–Popov–Terent'ev rate formula used to model depletion of the outer electron when estimating the ionization regime.","marker":"[35]"},{"why":"Imaginary-time-method formalism and Coulomb correction underlying the analytic action derivation.","marker":"[39]"},{"why":"Imaginary-time and matching procedures used to regularize the Coulomb action near the nucleus.","marker":"[40]"},{"why":"Saddle-point and barrier-suppression-field analysis for the symmetric two-electron configuration that sets the tunneling criterion.","marker":"[44]"},{"why":"Smirnov–Chibisov quasistatic rate used for the outer-electron ionization rate in the pulse-duration estimate.","marker":"[47]"}],"fun_headline_variants":["Electron pairs tunnel out together in strong fields","Collective tunneling: two electrons move as one","Side-by-side electron tunneling leaves momenta shoulders","Theory: joint tunneling for nonsequential double ionization"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The derivation assumes the most probable escape path is exactly the symmetric configuration x1 = x2, y1 = −y2, with the electron–electron repulsion term dominating the nuclear attraction inside the barrier; for regions where |y| ≥ |x|, no analytic correction is derived.","fun_headline_variants_meta":{"raw":{"variants":["Electron pairs tunnel out together in strong fields","Collective tunneling: two electrons move as one","Side-by-side electron tunneling leaves momenta shoulders","Theory: joint tunneling for nonsequential double ionization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000238,"raw_usage":{"total_tokens":1313,"prompt_tokens":677,"completion_tokens":636,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":421,"completion_tokens_details":{"reasoning_tokens":577}},"tokens_in":421,"tokens_out":636,"duration_ms":7461,"temperature":1.0,"reasoning_tokens":577,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T04:20:11.714742+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure two-electron coincidence lateral momenta for nonsequential double ionization of xenon in a short circularly polarized pulse with recollisions suppressed. If the collective channel exists, the p1y = −p2y diagonal should show side peaks or shoulders near p_y = (π^2 E^2/(16 Ip))^(1/6); a distribution identical to the non-interacting sequential background at the same field would refute the prediction. Additionally, at fields E much below the barrier-suppression field, a direct numerical evaluation of Im S along the full two-particle trajectory should reproduce Eq. (48); a mismatch there wo","supporting_citations":[{"cited_title":"Weber, M","cited_arxiv_id":null,"evidence_quote":"Original quasiparticle picture of collective tunneling whose rate the paper corrects by including electron–electron repulsion."},{"cited_title":"Moshammer, B","cited_arxiv_id":null,"evidence_quote":"Early collective-tunneling model that the paper shows is inconsistent with numerical calculations when repulsion is discarded."},{"cited_title":"Eichmann, M","cited_arxiv_id":null,"evidence_quote":"Previous numerical TDSE study that identified the collective flux in short pulses and supplies the non-interacting-electron comparison distributions."},{"cited_title":"Klaiber, K","cited_arxiv_id":null,"evidence_quote":"Perelomov–Popov–Terent'ev rate formula used to model depletion of the outer electron when estimating the ionization regime."},{"cited_title":"K¨ ubel, N","cited_arxiv_id":null,"evidence_quote":"Imaginary-time and matching procedures used to regularize the Coulomb action near the nucleus."},{"cited_title":"Breakdown of sequential tunnel ionization in ultrashort electromagnetic pulses","cited_arxiv_id":"2504.20583","evidence_quote":"Saddle-point and barrier-suppression-field analysis for the symmetric two-electron configuration that sets the tunneling criterion."},{"cited_title":"Perelomov, V","cited_arxiv_id":null,"evidence_quote":"Smirnov–Chibisov quasistatic rate used for the outer-electron ionization rate in the pulse-duration estimate."}],"review_version":1}