{"id":"8a7b69fb-e8de-480d-bbeb-6cd7be1451e5","arxiv_id":"2608.06647","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":8,"one_line_summary":"The electron g-factor in an InGaAs quantum dot molecule steps from -0.336 +/- 0.008 to -0.389 +/- 0.003 as the electron tunnels between the two dots, while the hole g-factor stays near 0.094 +/- 0.007.","lead":"Researchers tracked how the electron and hole g-factors of a charged exciton in a pair of coupled quantum dots respond as voltage pushes the electron from one dot to the other. This gives a spectroscopic way to tune each dot's spin splitting by an applied electric field, which matters for generating entangled multi-photon cluster states.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Predicted electron g-factor step is 0.026 versus measured 0.053, and the hole g-factor has the wrong trend; 'quantitatively reproduced' overstates the theory, though the experimental step is solid.","rationale":"The reader correctly identifies the unmeasured morphology as the main source of uncertainty in the theory-experiment comparison. My stress-test agrees with that, but finds a more immediate, internal problem: even using the parameters stated in the paper, the model does not quantitatively reproduce the headline electron g-factor step. The predicted step is 0.026, while the measured step is 0.053 ± 0.009, a discrepancy of about a factor of two and roughly 3σ. The paper itself concedes that the hole g-factor is both too small and has the wrong voltage trend. Therefore the abstract's 'quantitatively reproduced' is stronger than the evidence supports. This does not undermine the central experimental finding: the voltage-controlled step in g_e near the electron anticrossing is a robust, clean spectroscopic observation, with no observed magnetic-field dependence, and it does not rely on the k.p model. The model's partial success in reproducing the PLV energy levels, avoided-crossing widths, and the sign/position of the g_e step is real supporting evidence, but it is not enough to justify a fully quantitative claim. Since the reader already assigned a CONDITIONAL verdict, my concern sharpens the reasons for that conditionality without changing the verdict.","tokens_in":28,"tokens_out":6478,"duration_ms":116653,"concrete_test":"Perform a parameter-sensitivity study with the eight-band k.p code: vary (i) the Al0.33Ga0.67As barrier position from the current +a/2 offset to 0 and -a/2 relative to the nominal growth position, (ii) In content Ct and Cb within ±0.05, and (iii) z0 within ±1 nm, and compute the predicted g_e step and its voltage position. If no combination within independent growth tolerances yields a step within 2σ of 0.053 ± 0.009 while also reproducing the measured anticrossing energies (Δac ≈ 2.95 meV, singlet-triplet 0.655 meV, two-electron 0.676 meV), then the quantitative-reproduction claim fails. If such a set exists, the morphology concern is largely resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section III C reports calculated electron g-factors for the upper and lower dots of -0.424 and -0.398, respectively; the predicted step is therefore |Δg_e| = 0.026. The experimental step in Sec. II B is |Δg_e| = |-0.389 - (-0.336)| = 0.053 ± 0.009, a factor-of-two and roughly 3σ discrepancy. Appendix 3b similarly reports hole g-factors of 0.037-0.061 versus measured 0.078-0.110, with the calculated field dependence opposite to the data. Thus the abstract claim that 'our results are quantitatively reproduced by an eight-band k.p model' is not supported by the paper's own numbers. Because the step magnitude is the main quantity the theory is supposed to explain, the model agreement is at best semi-quantitative. This weakness is compounded by the reader's point that the morphology (truncated Gaussians, trumpet In profile, 2.5 nm AlGaAs barrier, and the ad hoc half-lattice-constant barrier offset toward the lower dot) is inherited from Refs. [20,44] and not independently measured for this device. The clean experimental observation of a voltage-tunable g_e step and its association with Ve- is not invalidated by this, since it comes directly from polarization-resolved PL with no B-field dependence; what is load-bearing is the 'quantitative reproduction' and the specific morphological interpretation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports polarization-resolved magneto-photoluminescence measurements of the electron and hole g-factors of the negatively charged trion X^- in a single InGaAs quantum dot molecule as a function of gate voltage. The central experimental finding is a step-like change in the electron g-factor from g_e = -0.336 ± 0.008 to g_e = -0.389 ± 0.003 at the electron tunneling resonance, with no observable magnetic-field dependence, while the hole g-factor stays nearly constant with a small modulation. The authors compare these data with an eight-band k·p model combined with a configuration-interaction treatment, and the abstract and conclusions state that the results are quantitatively reproduced by the theory. The paper also discusses the role of the AlGaAs tunnel barrier and the dot morphology in engineering the g-factor step.","tokens_in":17138,"tokens_out":3170,"duration_ms":28383,"significance":"If the experimental step is robust, the work is significant: it demonstrates electrostatic control of the electron Zeeman splitting in a quantum dot molecule, which is relevant for suppressing g-factor-mismatch dephasing and for photonic cluster-state protocols. The experimental measurement appears careful: g-factors are extracted from polarization-resolved data at three magnetic fields, and the absence of B-field dependence supports the interpretation. The theoretical framework is standard and the code is evidently mature, with prior applications to similar samples. However, the central quantitative claim is not supported by the paper's own numbers: the calculated electron g-factor step is about half the measured step, and the hole g-factor is substantially underestimated with the wrong field dependence. The theoretical interpretation therefore remains semi-quantitative, and the significance of the paper as a demonstration of predictive modeling is weakened accordingly.","major_comments":[{"comment":"The abstract and conclusions claim that the results are 'quantitatively reproduced' by the eight-band k·p model, but the numbers in Sec. III C do not support this. The calculated electron g-factors for the upper and lower dots are -0.424 and -0.398, giving a step of 0.026, whereas the measured step is |-0.389 - (-0.336)| = 0.053 ± 0.009. The discrepancy is roughly a factor of two and exceeds the experimental uncertainty by about 3σ. The individual values also disagree by 0.035-0.06. Because the step magnitude is the key quantity the model is intended to explain, this should be described as semi-quantitative or qualitative agreement, not quantitative reproduction.","section":"Abstract and Sec. III C"},{"comment":"The hole g-factor agreement is demonstrably poor. The calculated values span 0.037-0.061, while the measured g_h ranges from 0.078 ± 0.007 to 0.110 ± 0.004. Moreover, the text admits that the calculated field dependence is opposite to the measured one: the theory increases with increasing electric field while the experiment decreases. Since the trion g-factor involves both carriers, this undermines the claim that the model captures the voltage dependence of the trion Zeeman splitting, and it should be discussed as a limitation rather than as 'reasonable agreement.'","section":"Sec. III C and Appendix 3b"},{"comment":"The quantitative comparison rests on morphological parameters that are not independently measured for this device. The truncated-Gaussian shapes, trumpet-shaped In profile, Al0.33Ga0.67As barrier thickness, and especially the ad hoc assumption that the barrier sits half a lattice constant closer to the lower dot are all taken from prior modeling of similar samples (Refs. [20,44]) or introduced for this paper. The sensitivity of the calculated g-factor step to these choices is not tested. The reader cannot assess whether the remaining factor-of-two discrepancy in the electron step and the wrong hole trend arise from the model inputs or from missing physics. A parameter-sensitivity analysis or a scan over the uncertain morphology would be needed to support a quantitative claim.","section":"Sec. III C and Appendix 3a"}],"minor_comments":[{"comment":"The formulas g_e = |g_H + g_V|/2 and g_h = |g_H - g_V|/2 use absolute values, but the reported electron g-factor is negative. The sign convention should be stated explicitly, for example by defining the sign of the Zeeman splitting relative to the polarization assignments.","section":"Sec. II B"},{"comment":"In Fig. 5(b), the labels 'g_e in the lower QD' and 'g_e in the upper QD' are not self-explanatory; the text explains that these are evaluated at F = -5 kV/cm and F = 25 kV/cm, respectively. The caption should state this to avoid confusion.","section":"Fig. 5 caption"},{"comment":"The sentence 'the CI space consists of four electron states (two s-like states for each spin projection in each QD) and two hole states (the s-like states in the upper QD)' is ambiguous about whether the two hole states are the two spin projections. Please clarify.","section":"Sec. III A"},{"comment":"Reference [6] contains a typo: 'Ecomplete quantum control' should be 'Complete quantum control.'","section":"Reference [6]"},{"comment":"The terms 'H' and 'V' for horizontal and vertical polarization are used in figures and text but not defined in the figure captions; please define them at first use.","section":"Fig. 1(c) and Sec. II A"}],"recommendation":"major_revision","confidential_remarks":"The experimental measurement is solid and publishable, but the theoretical claims need careful recalibration. The abstract's 'quantitatively reproduced' is contradicted by the paper's own numbers, and the hole g-factor trend is opposite to experiment. I would ask the authors to either substantially improve the theory (e.g., by varying the morphology parameters within physically plausible bounds) or to rewrite the claims as semi-quantitative and clearly separate the robust experimental observation from the model-dependent interpretation. This is within the scope of a major revision. The paper also relies heavily on parameters from earlier same-group papers; that is not inappropriate, but the lack of an independent structural characterization should be acknowledged more prominently."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Plain English: the paper gives the first separate electron and hole g-factor characterization of the X- trion across the full voltage range in a QDM. That is genuinely new and useful. The clean polarization-resolved data at three magnetic fields show a step-like electron g-factor change at the tunneling resonance, from -0.336 to -0.389, with no B-field dependence. That is a direct spectroscopic fingerprint of molecular orbital formation and is solid.\n\nThe paper also does a good job placing this in context, distinguishing itself from Doty et al. and Liu et al., who didn't separate electron and hole contributions.\n\nWhere it gets soft is the theory. The eight-band k.p model does reproduce the qualitative step, but the numbers don't support \"quantitatively reproduced\" in the abstract. The calculated electron step is 0.026, half the measured 0.053, roughly 3 sigma off. The hole g-factor has the wrong trend: theory increases with electric field while measured decreases, and the calculated values (0.037-0.061) are well below the measured (0.078-0.110). The paper acknowledges this in Sec. III C, but the abstract and conclusion overstate the agreement.\n\nThere is also a modeling concern: the QD morphology (truncated Gaussians, trumpet In profile, AlGaAs barrier position) comes from prior same-group papers [20,44] plus an ad hoc half-lattice-constant shift of the barrier toward the lower dot. None of this is independently measured for this device. If those inputs are off, the theory-experiment match could be coincidental. No code or data is shipped for the theory, so a referee can't easily test the sensitivity.\n\nThe central experimental result does not depend on the model. So I see the paper as a solid experimental contribution with an over-enthusiastic theoretical wrapper.\n\nWho benefits: experimentalists working on QDM spin control and cluster-state generation, and theorists who want a benchmark for g-factor calculations in coupled dots.\n\nIt deserves a serious referee. I'd send it out, but ask the authors to tone down the \"quantitatively reproduced\" claim and add error bars or caveats on the theoretical step and hole trend.","headline":"Solid new data on separately resolved electron/hole g-factors in a QDM, but the k.p model overclaims quantitative agreement.","tokens_in":17742,"tokens_out":2116,"would_cite":true,"duration_ms":19157,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["71.70.Ej","73.21.La","78.67.Hc"],"model":"deepseek-v4-flash","headline":"The electron g-factor of a negatively charged trion in an InGaAs quantum dot molecule jumps in a step-like manner at the tunneling resonance, from $-0.336$ to $-0.389$, a fingerprint of molecular-orbital formation that lets a gate voltage…","keywords":["quantum dot molecule","electron g-factor","hole g-factor","tunneling resonance","eight-band k·p model","trion","magneto-photoluminescence","voltage-controlled Zeeman splitting"],"falsifier":"Cross-sectional scanning transmission electron microscopy of this exact sample could check whether the AlGaAs barrier really sits half a lattice constant closer to the lower dot as assumed; a centered barrier would invalidate the predicted asymmetry in the dot-resolved g-factors and the step magnitude. A second, simpler test is to measure $g_\\mathrm{e}(V)$ in a device from the same growth run with the interdot AlGaAs barrier removed, where the model predicts a much smaller step.","tokens_in":16571,"feed_emoji":"🧲","tokens_out":7809,"duration_ms":61153,"temperature":0.7,"pith_summary":"The paper reports that in a single InGaAs quantum dot molecule, the electron g-factor of a negatively charged trion changes abruptly from $-0.336 \\pm 0.008$ to $-0.389 \\pm 0.003$ when a gate voltage tunes the electron across the tunneling resonance between the two dots. This step is a direct spectroscopic fingerprint of molecular-orbital formation: the electron wavefunction shifts its localization from the lower to the upper dot, sampling different local strain, composition, and barrier proximity. The hole g-factor stays nearly flat at about $0.094$ with only a weak modulation near the anticrossings, attributed to Coulomb-mediated deformation by the tunneling electron. Eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ theory reproduces the electron step quantitatively and shows the step size can be engineered through the Al content of the interdot barrier. If correct, the result makes electric-field control of trion g-factors a practical tool for independently setting the Zeeman splitting of each dot, relevant for suppressing spin dephasing in coupled-spin quantum photonic protocols.","feed_headline":"Voltage swings electron g-factor at a quantum-dot tunnel resonance","feed_subtitle":"Step-like change fingerprints molecular-orbital formation and paves the way to voltage-tuned spin splittings.","key_machinery":"The argument is carried by the $X^-$ trion singlet line $X_{S0}^-$ measured by polarization-resolved magneto-photoluminescence in Voigt geometry: the H- and V-polarized transition pairs split by the in-plane magnetic field yield the electron and hole g-factors through $g_\\mathrm{e} = |g_H + g_V|/2$ and $g_\\mathrm{h} = |g_H - g_V|/2$. The theoretical reproduction uses an eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ envelope-function model with the magnetic field included gauge-invariantly, combined with configuration-interaction treatment of the trion restricted to the $s$-shell single-particle manifold. The QD geometry is specified by truncated-Gaussian shapes, a trumpet-shaped indium composition profile, and an $\\mathrm{Al}_{0.33}\\mathrm{Ga}_{0.67}\\mathrm{As}$ interdot barrier; strain is computed by continuum elasticity with second-order piezoelectric terms. The key mechanism is that the electron wavefunction's tails penetrate the AlGaAs barrier, whose bulk AlAs g-factor is positive ($+1.52$), partially compensating the negative host g-factor; the lower dot sits closer to the barrier, giving it a less negative g-factor than the upper dot.","core_discovery":"The central discovery is that the electron Zeeman response of a quantum dot molecule can be switched by a static electric field: as the final-state electron of the $X^-$ trion tunnels from the upper to the lower dot at $V_e^-$, the measured electron g-factor shifts in a step-like way from $g_\\mathrm{e} = -0.336 \\pm 0.008$ to $g_\\mathrm{e} = -0.389 \\pm 0.003$, while the hole g-factor remains nearly constant. The authors attribute the step to the electron wavefunction moving between two dots with distinct heights, compositions, strain profiles, and distances to the $\\mathrm{Al}_{0.33}\\mathrm{Ga}_{0.67}\\mathrm{As}$ interdot barrier, whose positive AlAs contribution partially compensates the negative GaAs/InAs g-factor. An eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ model reproduces the step and predicts dot-resolved g-factors of $-0.424$ (upper dot) and $-0.398$ (lower dot), in reasonable agreement with experiment, and shows that the step magnitude grows with barrier Al content, demonstrating that the effect can be engineered.","pith_inferences":["A natural test is to grow a second QDM without the AlGaAs interdot barrier: the model predicts a much smaller $g_\\mathrm{e}$ step, so a clean comparison would isolate the barrier's role from the dot-height difference.","One could exploit the step as a fast in-situ calibration: a polarization-resolved PLV sweep already contains the full $g_\\mathrm{e}(V)$ curve, so the tunnel resonance can be located without separate transport measurements.","The voltage-dependent $g_\\mathrm{e}$ suggests pulsed-gate sequences could dynamically switch the effective Zeeman field seen by a spin, an implicit pathway toward fast spin-echo or decoupling protocols that the paper does not itself demonstrate.","Because the model assigns the step to wavefunction tails in the barrier, interface roughness or alloy fluctuations at the AlGaAs/InGaAs interfaces should smear or shift the step; measuring several nominally identical devices would quantify this sensitivity."],"forward_implications":["A single gate voltage can independently set the electron Zeeman splitting of each dot, allowing g-factor mismatches between tunnel-coupled dots to be compensated in situ, which suppresses pure dephasing in two-qubit spin gates.","The step in $g_\\mathrm{e}$ at $V_e^-$ provides an all-optical, polarization-resolved readout of where the electron wavefunction is localized, i.e., of the molecular-orbital character, across the full voltage range.","The magnitude of the g-factor step can be engineered by choosing the Al content of the interdot barrier: theory predicts the step grows progressively as $x$ increases from $0.11$ to $0.33$ in $\\mathrm{Al}_x\\mathrm{Ga}_{1-x}\\mathrm{As}$.","The hole g-factor is only weakly affected by the tunneling electron, so independent electron and hole Zeeman tuning is possible, which is important for double-lambda optical schemes.","No magnetic-field dependence of $g_\\mathrm{e}$ or $g_\\mathrm{h}$ is observed between 6 T and 8 T, meaning the extracted g-factors are robust field-independent parameters for spin control protocols."],"supporting_citations":[{"why":"Supplies the QDM growth and the truncated-Gaussian/trumpet-composition geometry that the eight-band model inherits; the paper's simulation parameters come from this reference.","marker":"[20]"},{"why":"Demonstrated electrically tunable g-factors in QDMs and identified wavefunction sampling of the tunnel barrier as the mechanism; the electron step reported here is the electron analogue of that hole-side effect.","marker":"[24]"},{"why":"Established dot-specific g-factors and the quantum-confined Stark shift in composition-engineered In(Ga)As dots, used to explain the weak voltage dependence of $g_\\mathrm{h}$.","marker":"[33]"},{"why":"Introduced an AlGaAs barrier into an InAs QDM and showed a 50% in situ change in exciton g-factor; the present work resolves the electron and hole contributions separately that this study left entangled.","marker":"[38]"},{"why":"Provided the eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ framework with gauge-invariant magnetic field incorporation that quantitatively reproduces resonant g-factor changes in QDMs.","marker":"[39]"},{"why":"Supplies the few-body Hamiltonian used to identify trion charge states and the QDM geometry/composition parameters adopted for this sample.","marker":"[44]"},{"why":"Provides the gauge-invariant discretization scheme used to include the magnetic field in the eight-band model.","marker":"[55]"},{"why":"Validates the eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ plus configuration-interaction approach for computing g-factors and optical spectra of self-assembled QDs.","marker":"[36]"},{"why":"Shows that g-factor mismatch between tunnel-coupled dots causes pure dephasing, the practical problem that voltage-tunable g-factors are meant to solve.","marker":"[28]"}],"fun_headline_variants":["Voltage step flips electron g-factor in quantum dot molecule","Tunneling resonance makes electron g-factor jump between dots","Electric field tunes electron g-factor at dot-molecule crossover","G-factor step at tunnel barrier enables voltage control of spin"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative match between theory and the measured $g_\\mathrm{e}$ step rests on the assumed dot geometry and composition profile (truncated Gaussian shapes, a trumpet-shaped indium profile, and an $\\mathrm{Al}_{0.33}\\mathrm{Ga}_{0.67}\\mathrm{As}$ barrier placed half a lattice constant closer to the lower dot), values inherited from earlier modeling rather than measured on this specific sample; if those morphological inputs are wrong, the agreement could be coincidental.","fun_headline_variants_meta":{"raw":{"variants":["Voltage step flips electron g-factor in quantum dot molecule","Tunneling resonance makes electron g-factor jump between dots","Electric field tunes electron g-factor at dot-molecule crossover","G-factor step at tunnel barrier enables voltage control of spin"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000244,"raw_usage":{"total_tokens":1615,"prompt_tokens":1108,"completion_tokens":507,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":724,"completion_tokens_details":{"reasoning_tokens":439}},"tokens_in":724,"tokens_out":507,"duration_ms":5027,"temperature":1.0,"reasoning_tokens":439,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T04:10:40.871707+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Cross-sectional scanning transmission electron microscopy of this exact sample could check whether the AlGaAs barrier really sits half a lattice constant closer to the lower dot as assumed; a centered barrier would invalidate the predicted asymmetry in the dot-resolved g-factors and the step magnitude. A second, simpler test is to measure $g_\\mathrm{e}(V)$ in a device from the same growth run with the interdot AlGaAs barrier removed, where the model predicts a much smaller step.","supporting_citations":[{"cited_title":"Lienhart, K","cited_arxiv_id":null,"evidence_quote":"Supplies the QDM growth and the truncated-Gaussian/trumpet-composition geometry that the eight-band model inherits; the paper's simulation parameters come from this reference."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrated electrically tunable g-factors in QDMs and identified wavefunction sampling of the tunnel barrier as the mechanism; the electron step reported here is the electron analogue of that hole-side effect."},{"cited_title":"Nakaoka, T","cited_arxiv_id":null,"evidence_quote":"Established dot-specific g-factors and the quantum-confined Stark shift in composition-engineered In(Ga)As dots, used to explain the weak voltage dependence of $g_\\mathrm{h}$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduced an AlGaAs barrier into an InAs QDM and showed a 50% in situ change in exciton g-factor; the present work resolves the electron and hole contributions separately that this study left entangled."},{"cited_title":"Medeiros-Ribeiro, E","cited_arxiv_id":null,"evidence_quote":"Provided the eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ framework with gauge-invariant magnetic field incorporation that quantitatively reproduces resonant g-factor changes in QDMs."},{"cited_title":"Thalacker, M","cited_arxiv_id":null,"evidence_quote":"Supplies the few-body Hamiltonian used to identify trion charge states and the QDM geometry/composition parameters adopted for this sample."},{"cited_title":"Bester, X","cited_arxiv_id":null,"evidence_quote":"Provides the gauge-invariant discretization scheme used to include the magnetic field in the eight-band model."},{"cited_title":"Gawarecki, A","cited_arxiv_id":null,"evidence_quote":"Validates the eight-band $\\mathbf{k}{\\cdot}\\mathbf{p}$ plus configuration-interaction approach for computing g-factors and optical spectra of self-assembled QDs."},{"cited_title":"Gawełczyk, M","cited_arxiv_id":null,"evidence_quote":"Shows that g-factor mismatch between tunnel-coupled dots causes pure dephasing, the practical problem that voltage-tunable g-factors are meant to solve."}],"review_version":1}