{"id":"3d8f450f-92ad-4c49-9ce9-5b4d9bf7e148","arxiv_id":"2507.23352","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Simulations of a suspended silica rib waveguide show two-color evanescent Rb traps with lateral depths up to 317.6 µK (mono-resonant) and 313.3 µK (dual-resonant), exceeding 0.3 mK.","lead":"This modeling paper shows that trapping rubidium atoms above a tiny suspended glass waveguide with two colors of laser light gets much deeper when the higher-frequency color is carried by two waveguide modes instead of one, and when the two colors are tuned near two different atomic transitions. It predicts trap depths above 0.3 mK, which would make planar photonic atom chips more practical for quantum sensing and computing.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Coherent interference between TE00 and TE01 is neglected; the reported lateral depths assume an incoherent intensity sum, but co-propagating modes from a single laser have a fixed relative phase and the trap depth may be reduced.","rationale":"The reader identified the same weak point. I agree it is the most load-bearing: the paper's reported depths are computed from the incoherent intensity sum, but the proposed waveguide excitation (single-frequency blue light in two TE modes) is naturally coherent. The interference term is not tiny: for equal modal powers, its magnitude is comparable to the individual intensities at the TE01 maxima. The odd symmetry of the cross term means the two lateral barriers are unequal, and the shallower one determines the trap depth. Since the relative phase advances with z, the barrier height also modulates along the guide; atoms that are trapped in 2D can escape over the weakest point. The paper's own Fig. 4 shows the sum of intensities, not the field. A calculation with a fixed phase is needed. If the interference seriously degrades the depth, the design still may work with phase-controlled excitation or with two frequencies for the two modes, but the current manuscript does not demonstrate that. I do not see a more fundamental flaw: the mode parameters are plausible, the potential calculation is internally consistent, and the dual-resonant scattering rates are within the stated approximations. The lack of reproducibility (mode profiles not tabulated) is a secondary concern. Hence the verdict remains CONDITIONAL, and the reader's verdict is unchanged.","tokens_in":10001,"tokens_out":10072,"duration_ms":78934,"concrete_test":"Using the mode profiles and propagation constants reported in Sections 5-6, compute the coherent blue intensity I_b(z,y) = |c0 E00(y) exp(iβ00 z) + c1 E01(y) exp(iβ01 z)|² with c0, c1 set to the stated mode powers. For each z over one beat period, evaluate the total potential (Eq. 10) and the lateral depth ΔU(z) between the potential minimum and the lateral saddle point. Report the minimum of ΔU(z) for the 698/1300 nm and 420.2/980 nm cases. If that minimum is below the corresponding single-TE00 depth (99 µK / 44.8 µK), the two-mode scheme's advantage disappears; if it remains above ~100 µK, the concern is mitigated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Central to the claimed trap depths is the representation of the blue-detuned field as the sum of TE00 and TE01 intensities (dotted curve, Fig. 4). This is an incoherent addition. In practice, the two co-propagating TE modes are excited with a fixed relative phase, so the intensity is I_b(z,y) = I00(y) + I01(y) + 2√(I00 I01) cos(Δβ z + φ) times the sign of E00 E01. Because E00 is even in y and E01 is odd, the cross term is odd, making the blue potential asymmetric. On one side the repulsive barrier is reduced; the lateral depth is set by that side. Moreover, the term oscillates with z with period 2π/Δβ ≈ 50 µm (698 nm) and ≈ 41 µm (420.2 nm), so the effective trap depth for atoms moving along the guide is the minimum over z, not the z-average computed in Sections 5-6. The paper does not propose a mechanism for incoherent excitation (e.g., separate input beams with a frequency offset large compared to atomic linewidths), so the reported 317.6 µK and 313.3 µK are not the physical trap depth for a single-laser implementation. This is the load-bearing assumption: all headline numbers in Sections 5 and 6 depend on it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper models two-colour evanescent dipole traps for 87Rb above suspended silica rib waveguides. Its central proposal is to guide the blue-detuned frequency component in the two lowest transverse modes (TE00 and TE01) while the red-detuned component propagates in a single TE00 mode, thereby broadening the lateral intensity profile of the blue field. Applying this to a 1300/698 nm mono-resonant design and to a 980/420.2 nm dual-resonant design, the paper reports minimum lateral trap depths of 317.6 µK and 313.3 µK, respectively, and states that total depths of 0.3 mK are reachable under reasonable conditions. The analytic framework of Section 2 is internally consistent, and the quoted η, S, and penetration-depth values are mutually consistent with the reported mode simulations.","tokens_in":10254,"tokens_out":10436,"duration_ms":119038,"significance":"If the central claim is correct, the two-mode strategy would directly address the well-known lateral-anisotropy problem of planar waveguide evanescent traps, a problem that has so far prevented experimental realization. The paper is a forward calculation with no fitted parameters: the trap depths follow from specified geometries, mode powers, and literature atomic matrix elements, and the paper explicitly checks η and S against the penetration depths. It also provides scattering rates and heating estimates for the proposed configurations, which is valuable for experimental planning. However, the headline depth values rest on an unphysical incoherent superposition of the two blue modes, so the quantitative conclusions are not yet established.","major_comments":[{"comment":"The statement in Section 7 that mode-selective coupling can be achieved with directional couplers does not resolve the coherence issue. Directional couplers set the amplitudes of the TE00 and TE01 modes, but the relative phase at the trapping region depends on the coupler design and the propagation length to that region. Unless the two modes are driven at distinctly different frequencies, the interference term remains and the trap depth is z-dependent. The manuscript should either provide a specific phase-controlled excitation scheme or explicitly demonstrate that the resulting interference pattern does not reduce the lateral barrier below the quoted values.","section":"Sec. 5, Fig. 4; Sec. 6, Fig. 9"}],"minor_comments":[{"comment":"The caption for Fig. 6 incorrectly states the wavelengths as λr = 782 nm and λb = 778 nm; this appears to be a copy-paste from the Fig. 2 caption. The figure shows the 1300/698 nm two-mode case and the caption should state those wavelengths.","section":"Fig. 6 caption"},{"comment":"The sentence 'The frequency of the \"red\" component is chosen to be far below the 5S−5P transition at λb = 980nm' uses the subscript b for the 980 nm mode; this should be λr = 980 nm.","section":"Sec. 6, text near Eq. (6)"},{"comment":"The numerical simulations of the waveguide modes are attributed to reference [21] in Section 5, whereas Sections 4 and 6 attribute the same kind of simulation to reference [22]. The reference number in Section 5 appears to be a typographical error and should be corrected to [22].","section":"Sec. 5, first paragraph"},{"comment":"Fig. 4 is described as showing the sum of the TE00 and TE01 intensities 'with the same amplitudes', while the actual trap calculation in Fig. 6 uses powers of 7.73 mW and 8.65 mW for the two modes. Please clarify whether the dotted curve is only illustrative or whether it corresponds to the exact power ratio used in the trap-depth calculation.","section":"Fig. 4 and Sec. 5, paragraph before Fig. 6"},{"comment":"The van der Waals term is written as '-C3λef f/2π / x3(x + λef f/2π)', which is hard to parse. Please write the expression with explicit brackets, e.g., -C3 (λeff/2π) / [x^3 (x + λeff/2π)], to avoid ambiguity.","section":"Eq. (10)"}],"recommendation":"major_revision","confidential_remarks":"The coherence issue is genuinely load-bearing and not a stylistic point: every quantitative result in Sections 5 and 6 assumes that the TE00 and TE01 intensities add without interference. The author should be asked to recompute the coherent potential and quote the minimum barrier over z, or else to provide a concrete excitation scheme that makes the modes incoherent. The rest of the manuscript's analytic framework and numerical consistency checks are sound, so I would not reject the paper at this stage; the design idea is worth pursuing once the coherent case is quantified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this paper is worth reading for its two-mode idea, but the headline depths are not yet physically grounded. The author proposes guiding the upper-frequency component in both the TE00 and TE01 modes of a suspended rib waveguide, and in the dual-resonant version tuning the two colors to different Rb transitions. That combination is new in the planar-waveguide literature I know. The analytic machinery is clean—I checked Eq. (5) and the quoted efficiency coefficients, and they are consistent. The forward calculation from known Rb matrix elements is honest, with no fitting to a target. The plotted potentials match the stated mode parameters and powers. If the underlying assumption is valid, the result would matter: it could push all-optical planar atom traps from microkelvin depths to a few hundred microkelvins.\n\nThe soft spot is exactly where the stress-test points. The dotted curve in Fig. 4 is an incoherent sum of TE00 and TE01 intensities. If both modes are excited by the same 698 nm light, the physical field is a coherent superposition, and the interference term 2 Re[c0 c1* E00 E01*] cos(Δβ z) is odd in y and oscillates along z with period roughly 50 µm. The effective lateral trap depth is then set by the weaker side and the minimum over z, not the z-averaged field. Sections 5 and 6 never mention this. It is a load-bearing omission, though a curable one: one could excite the two modes with separate frequencies offset beyond the atomic response bandwidth, or with separately phase-controlled couplers. The paper does not propose or discuss either.\n\nOther concerns are smaller. The mode-solver input is not fully reproducible from the manuscript, and the numbers have no error bars. The dual-resonant 420.2 nm example sits close to resonance, so the 298 s-1 scattering rate and the 2 s lifetime claim are sensitive to detuning and intensity. The relation to Ref. [13], which already uses high-order transverse modes for two-color atom guiding, could be drawn more sharply.\n\nIn short: a transparent design study with a real new idea and one serious unaddressed physics issue. It deserves a referee, but not a quick accept. I would send it to peer review and ask the author to either model the coherent interference or specify an incoherent excitation scheme, and to provide enough mode-solver detail to reproduce the figures.","headline":"Genuinely new two-mode design idea with internally consistent numbers, but the headline trap depths rest on an unjustified incoherent-addition assumption for the two blue modes.","tokens_in":10818,"tokens_out":3924,"would_cite":true,"duration_ms":54428,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A two-mode blue field in a suspended rib waveguide can trap rubidium atoms at depths above 0.3 mK.","keywords":["two-colour evanescent trap","optical rib waveguide","87Rb","TE01 mode","optical dipole potential","atom chip","suspended waveguide","integrated atomo-photonics"],"falsifier":"Compute the full three-dimensional optical field with both blue modes, $E=c_0 E_{00}(x,y)e^{i\\beta_{00}z}+c_1 E_{01}(x,y)e^{i\\beta_{01}z}$, and include the interference cross term $2\\,\\mathrm{Re}[c_0 c_1^* E_{00}E_{01}^*]\\cos((\\beta_{00}-\\beta_{01})z)$. If the resulting longitudinal corrugation, with period $2\\pi/(\\beta_{00}-\\beta_{01})\\approx 50\\,\\mu\\mathrm{m}$ at 698 nm, lowers the minimum trap depth below the quoted values, then the reported depths are z-averaged rather than the field an atom experiences. A measurement of atom survival versus position along the guide would settle which description is correct.","tokens_in":9767,"feed_emoji":"⚛️","tokens_out":16809,"duration_ms":162433,"temperature":0.7,"pith_summary":"This paper argues that the lateral depth of a two-colour evanescent atom trap on a suspended silica rib waveguide can be increased dramatically by guiding the blue-detuned light in the two lowest transverse modes instead of one. In the 1300/698 nm mono-resonant design, adding the TE01 mode to the blue field raises the minimum lateral trap depth from about 99 µK to 317.6 µK; in the dual-resonant 980/420.2 nm design, tuned to two different rubidium transitions, the same technique gives 313.3 µK. The paper treats the two blue modes as adding in intensity. If the claim holds, planar photonic waveguides become practical platforms for coherent manipulation of ultracold atoms, with applications in quantum sensing, clocks, and neutral-atom quantum computing.","feed_headline":"Two-mode blue light deepens a planar waveguide atom trap to 0.3 mK","feed_subtitle":"Adding the second guided mode at 698 nm lifts the lateral trap depth from 99 to 318 µK for rubidium.","key_machinery":"The central object is the bi-exponential evanescent dipole potential $U_{\\mathrm{dip}}(x)=U_r e^{-2x/d_r}+U_b e^{-2x/d_b}$, where the red-detuned component is attractive ($U_r<0$) and the blue-detuned component is repulsive ($U_b>0$), with penetration depths $d_r>d_b$. Laterally, stable trapping requires the red field to dominate near the trap minimum. The paper's lever is to guide the blue light in the two lowest transverse modes, TE00 and TE01; because the TE01 intensity peaks away from the centre, adding it broadens the blue repulsive profile and deepens the lateral well. The efficiency coefficient $\\eta=(d_r-d_b)/d_r$ quantifies how much of the red potential is usable at the minimum, and the paper reports values 0.644 and 0.74 for the two designs.","core_discovery":"The central claim is that a two-mode suspended optical rib waveguide, in which the higher-frequency blue-detuned component propagates in the TE00 and TE01 transverse modes while the lower-frequency red-detuned component propagates in a single TE00 mode, produces a two-colour evanescent dipole trap for 87Rb with a lateral depth of about 0.3 mK or more. The mechanism is that the TE01 blue mode has intensity maxima away from the waveguide centre, so the combined blue repulsive profile is broader and lets the red attractive mode dominate at the trap minimum. For $\\lambda_r = 1300$ nm and $\\lambda_b = 698$ nm the paper computes a minimum lateral trap depth of 317.6 µK at milliwatt-level powers, with spontaneous scattering rates of a few s−1 and a heating rate near 0.8 µK/s. For the dual-resonant design, with $\\lambda_r = 980$ nm and $\\lambda_b = 420.2$ nm, it computes 313.3 µK, with a scattering rate of 298 s−1 and an estimated trapping time of about 2 s. The two designs are characterized by efficiency coefficients of 0.644 and 0.74, respectively.","pith_inferences":["If coherent beating between the two blue modes is not suppressed, the trap potential is corrugated along the waveguide axis with a period on the order of 50 µm at 698 nm, so the quoted depths describe an intensity-averaged field rather than the field an atom moving along the guide would see.","The two-mode broadening recipe could be transferred to other alkali species and other waveguide cross-sections, and possibly applied with the roles of red and blue light reversed to shape the attractive rather than the repulsive potential.","The dual-resonant design's scattering rate of 298 s−1 at 420.2 nm makes it most suitable for short interrogation cycles, while the far-detuned 1300/698 nm design is better suited to long-coherence storage; a hybrid sequence could exploit both.","A direct experimental test would load atoms from an optical tweezer into the waveguide trap and measure the lifetime and heating rate; matching the predicted 2 s at 420.2 nm would validate the scattering calculation."],"forward_implications":["For the 1300/698 nm waveguide, the minimum lateral trap depth rises from 99 µK without the TE01 mode to 317.6 µK with it, at similar milliwatt-level powers and with negligible heating of roughly 0.8 µK/s.","The dual-resonant 980/420.2 nm design reaches 313.3 µK, and its shorter blue wavelength is useful for building optical lattices and Bragg splitting at waveguide intersections.","The dual-resonant configuration can be switched to single-mode operation by replacing the 420 nm double mode with a single 640 nm mode, which may improve stability at waveguide crossings for atom interferometers.","The approach is not specific to rubidium; the paper states it applies to many other types of atoms as well.","The gain from the TE01 mode saturates at high relative intensities, so a modest TE01 power provides most of the lateral depth increase."],"supporting_citations":[{"why":"Establishes the two-colour evanescent light trap concept that this work extends to rib waveguides.","marker":"[5]"},{"why":"Shows suspended membrane waveguides give evanescent penetration depths comparable to nanofibres, motivating the suspended rib design.","marker":"[10]"},{"why":"Supplies the optical dipole potential formalism and ac polarizability definitions used for all potential calculations.","marker":"[17]"},{"why":"Provides relativistic many-body rubidium polarizability data used to compute dipole potentials and scattering rates.","marker":"[18]"},{"why":"Supplies critically evaluated 87Rb energies and dipole matrix elements for the 5P transitions.","marker":"[19]"},{"why":"Supplies measured dipole matrix elements for the 6P transitions used in the dual-resonant design.","marker":"[20]"},{"why":"Provides the van der Waals coefficient and effective wavelength for the Rb-SiO2 surface interaction term.","marker":"[21]"},{"why":"Provides the numerical mode solver used to obtain the TE00/TE01 field distributions and propagation constants.","marker":"[22]"},{"why":"Documents decay channels of the 6P state used to estimate the spontaneous scattering rate at 420 nm.","marker":"[23]"}],"fun_headline_variants":["Two-mode photonic waveguide trap reaches 0.3 mK","Dual-resonant rib waveguide traps atoms at 0.3 mK","2-mode planar guide lifts atom trap depth to 318 µK","Photonic waveguide combo deepens atom trap to 0.3 mK"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quoted trap depths assume the two blue-detuned guided modes add as intensities; if they instead interfere coherently, the potential along the waveguide is corrugated and the depth an atom actually experiences may be lower.","fun_headline_variants_meta":{"raw":{"variants":["Two-mode photonic waveguide trap reaches 0.3 mK","Dual-resonant rib waveguide traps atoms at 0.3 mK","2-mode planar guide lifts atom trap depth to 318 µK","Photonic waveguide combo deepens atom trap to 0.3 mK"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000678,"raw_usage":{"total_tokens":3076,"prompt_tokens":936,"completion_tokens":2140,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":552,"completion_tokens_details":{"reasoning_tokens":2060}},"tokens_in":552,"tokens_out":2140,"duration_ms":18007,"temperature":1.0,"reasoning_tokens":2060,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T10:50:07.184663+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full three-dimensional optical field with both blue modes, $E=c_0 E_{00}(x,y)e^{i\\beta_{00}z}+c_1 E_{01}(x,y)e^{i\\beta_{01}z}$, and include the interference cross term $2\\,\\mathrm{Re}[c_0 c_1^* E_{00}E_{01}^*]\\cos((\\beta_{00}-\\beta_{01})z)$. If the resulting longitudinal corrugation, with period $2\\pi/(\\beta_{00}-\\beta_{01})\\approx 50\\,\\mu\\mathrm{m}$ at 698 nm, lowers the minimum trap depth below the quoted values, then the reported depths are z-averaged rather than the field an atom experiences. A measurement of atom survival versus position along the guide would settle which description is correct.","supporting_citations":[{"cited_title":"Relativistic many-body calculations of electric- dipole matrix elements, lifetimes, and polarizabilities in rubidium,","cited_arxiv_id":null,"evidence_quote":"Provides relativistic many-body rubidium polarizability data used to compute dipole potentials and scattering rates."},{"cited_title":"An atomic trap based on evanescent light waves,","cited_arxiv_id":null,"evidence_quote":"Establishes the two-colour evanescent light trap concept that this work extends to rib waveguides."},{"cited_title":"Characterization of suspended membrane waveguides towards a photonic atom trap integrated platform,","cited_arxiv_id":null,"evidence_quote":"Shows suspended membrane waveguides give evanescent penetration depths comparable to nanofibres, motivating the suspended rib design."},{"cited_title":"Optical dipole traps for neutral atoms,","cited_arxiv_id":null,"evidence_quote":"Supplies the optical dipole potential formalism and ac polarizability definitions used for all potential calculations."},{"cited_title":"Critically evaluated theoretical energies, lifetimes, hyperfine constants, and multipole polarizabilities in 87Rb,","cited_arxiv_id":null,"evidence_quote":"Supplies critically evaluated 87Rb energies and dipole matrix elements for the 5P transitions."},{"cited_title":"Precision measurement of transition matrix elements via light shift cancellation,","cited_arxiv_id":null,"evidence_quote":"Supplies measured dipole matrix elements for the 6P transitions used in the dual-resonant design."},{"cited_title":"Coupling a single trapped atom to a nanoscale optical cavity,","cited_arxiv_id":null,"evidence_quote":"Provides the van der Waals coefficient and effective wavelength for the Rb-SiO2 surface interaction term."},{"cited_title":"Wave-matching mode solver for rectangular dielectric optical waveguides,","cited_arxiv_id":null,"evidence_quote":"Provides the numerical mode solver used to obtain the TE00/TE01 field distributions and propagation constants."},{"cited_title":"Narrow-line cooling of87Rb using 5S1/2→ 6P1/2 open transition at 420 nm,","cited_arxiv_id":null,"evidence_quote":"Documents decay channels of the 6P state used to estimate the spontaneous scattering rate at 420 nm."}],"review_version":1}