{"id":"81eda976-5e07-4822-8947-1f2c41939d03","arxiv_id":"2508.18639","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Circularly polarized light deterministically trains, directly switches, and spatially patterns the ferromagnetic polarization and thus the integer and fractional Chern insulator states in twisted MoTe2.","lead":"This paper shows that circularly polarized light can flip the magnetic order that controls topological states in twisted bilayer MoTe2, switching both integer and fractional Chern insulator phases. It also writes and erases magnetic domains, suggesting a route to reprogrammable topological circuits and quantum memory.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Optical control of Chern number is inferred solely from trion-PL helicity; without transport or another topological probe, a helicity flip does not by itself prove the Chern number flipped.","rationale":"The reader's weakest assumption is exactly the link from PL helicity to Chern number, and I agree that this is the soft spot. The paper's data are internally consistent, and the helicity-selective training, direct switching, dynamic modulation, and spatial patterning are compelling as optical control of ferromagnetic polarization. Prior work supports interpreting trion-PL helicity as a readout of ferromagnetic polarization, and the theoretical relation between valley polarization and Chern number in twisted MoTe2 is standard, so there is no internal contradiction or reason to doubt the basic observation. However, the title and abstract claim optical control of the CI and FCI states themselves, while the switched state is read out solely by PL. That makes the central claim one inference beyond the data: a spin/valley population effect or a local domain flip under the beam spot would also produce the observed helicity reversal without a bulk Chern-number switch. A transport measurement of the quantized Hall conductance before and after optical pumping is the natural arbiter, and it is a concrete, feasible check on the same device platform. Pending that check, conditional acceptance is appropriate: the optical control of ferromagnetism stands, while the Chern-number conclusion should be either verified by transport or explicitly qualified as an inference. This is a request for evidence matching the strength of the claim, not a rejection of the experimental work.","tokens_in":12311,"tokens_out":7582,"duration_ms":75538,"concrete_test":"Using the same device geometry (or a co-fabricated transport device from the same crystal), perform four-terminal transport at B = 0 while monitoring PL helicity. At ν = -1 and ν = -2/3, after initialization, measure R_xy and R_xx; then apply the same σ± trion-resonant pump used in Fig. 3, turn it off, and re-measure R_xy, R_xx, and PL helicity. The concern is settled if and only if R_xy changes sign and remains quantized (±h/e^2 at ν = -1; ±3h/(2e^2) at ν = -2/3) with vanishing R_xx, in registry with the PL helicity flip. If R_xy does not flip or loses quantization, the paper should be revised to claim optical switching of ferromagnetic polarization rather than of Chern number.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is that circularly polarized trion-resonant pumping flips the sign of the Chern number of the CI (ν = -1) and FCI (ν = -2/3) states. The only experimental evidence for this is the reversal of trion-PL helicity before versus after pumping (Figs. 1d, 1e, 3b, and 4), together with the previously established correlation between PL helicity and ferromagnetic polarization. The argument requires two links: (i) PL helicity faithfully encodes ferromagnetic polarization, and (ii) ferromagnetic polarization uniquely determines the Chern number sign. Link (i) has independent support from prior reports and from the RMCD phase map, but link (ii) is assumed from theory and is not verified in the optically switched state. A helicity-selective metastable valley-polarized hole population under the laser spot, or a local ferromagnetic-domain reversal that does not propagate through the bulk, would also produce the observed PL flip. The manuscript's inserted Fig. 3 caption even states that ρ behaves similarly in Chern-insulator phases and in the CFL region near ν = -1/2, so a near-unity ρ after pumping does not by itself certify a Chern insulator. The experiments are a convincing demonstration of optical control of moiré ferromagnetism; the stronger statement that the integer and fractional Chern numbers have been switched is an inference that needs a direct topological probe.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports helicity-selective optical control of ferromagnetic polarization in twisted bilayer MoTe2 at zero magnetic field, using trion-resonant circularly polarized pumping. At low power, optical training during gate sweeps prepares either polarization at fillings ν=-1 and -2/3; at higher power, direct pumping switches polarization at fixed filling. The authors infer flipping of the Chern number of the integer and fractional Chern insulators from the near-unity reversal of trion-PL helicity. They also demonstrate dynamic helicity modulation and spatially resolved optical writing of ferromagnetic domains, and attribute the effect's filling and electric-field dependence to a gap-enhanced optical torque from valley-polarized holes.","tokens_in":12568,"tokens_out":6943,"duration_ms":66821,"significance":"If the inference is accepted, this is an important advance: it demonstrates all-optical control of the magnetic order underlying a zero-field fractional Chern insulator, with potential for programmable topological domains and memories. The experimental data are internally consistent and well controlled: switching is reproducible, occurs only near the CI/FCI fillings, has threshold and electric-field dependence consistent with the gap-enhanced torque picture, and is visualized in spatial maps. The manuscript also states that source data are provided to reproduce the plots, which strengthens verifiability. The main limitation is interpretive: the Chern number itself is not measured after switching; the conclusion rests on the established but not device-specific calibration linking PL helicity to ferromagnetic polarization, and on the theoretical relation between that polarization and the Chern-number sign.","major_comments":[{"comment":"The conclusion that the Chern number of the ν=-1 CI and ν=-2/3 FCI states has been flipped rests entirely on the reversal of trion-PL helicity. The manuscript does not measure a topological response (quantized Hall conductance, Hall voltage, or chiral edge transport) in the optically switched state. Since PL helicity is a proxy for valley/magnetic polarization rather than a direct measure of C, the observed flip is also compatible with a laser-induced local domain reversal or with a metastable valley-polarized hole population; the spatial maps in Fig. 4c and Extended Data Fig. 8 show a written polarization domain but do not prove that it is a Chern-insulator domain. I recommend either adding a direct topological probe after switching or explicitly limiting the claims to optical control of ferromagnetic polarization, with the Chern-number switching stated as a theoretically grounded inference.","section":"Fig. 1d-e, Fig. 3b, and main-text claims"},{"comment":"A second caption numbered 'Fig. 3' (the 'Phase space comparison of ferromagnetism and trion photoluminescence helicity and temperature dependence' caption) appears in the manuscript. It explicitly states that ρ behaves similarly within the Chern-insulator phases and in the putative composite-Fermi-liquid region near ν=-1/2. This is directly relevant to the inference from ρ to topology: a near-unity ρ after pumping does not distinguish a Chern insulator from a CFL or another ferromagnetic state. The presence of this caption and its content underscores that PL-helicity readout alone is insufficient to certify Chern-number switching, and the duplication must be fixed editorially.","section":"Duplicated Fig. 3 caption and Fig. 4c"}],"minor_comments":[{"comment":"The text states that the memory effect of ferromagnetic orientation is 'consistent with prior reports16, 28', but Ref. 28 (Xiao et al., Interface engineering of quantum Hall effects in digital transition metal oxide heterostructures) does not appear to support this claim; please verify and correct the citation.","section":"References 16 and 28"},{"comment":"In the caption of Fig. 1, 'The PL in (d&e) are excited with linearly polarized HeNe laser' should be revised to 'The PL in (d,e) is excited with a linearly polarized HeNe laser'.","section":"Fig. 1 caption"},{"comment":"The quoted Curie temperatures (14 K near ν=-1 and 4.5 K near ν=-2/3) are not accompanied by a reference or an extraction method; please clarify their source.","section":"Main text, Curie temperatures"},{"comment":"The axis label D/ε_o appears without a definition; please define the displacement field in the main text or figure captions, and similarly ensure that the definition of ρ in the text is repeated consistently in the Methods or figure captions.","section":"Figure axes and Eq. (1)"}],"recommendation":"major_revision","confidential_remarks":"This is a high-impact experimental result from a well-established group, and the data are internally consistent. The main issue is the gap between the measured quantity (PL helicity) and the claimed quantity (Chern number), which is a load-bearing inference rather than a direct measurement. I would not reject, but I would ask the editor to require either a transport or local-edge probe of the switched state, or a substantial softening of the title and conclusions. The duplicated Fig. 3 caption and citation issues also indicate that the manuscript needs a careful editorial pass."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The genuinely new thing here is helicity-selective optical control of the ferromagnetic polarization in twisted MoTe2 at the integer (ν=-1) and fractional (ν=-2/3) Chern insulator states: low-power optical training, higher-power direct switching, dynamic helicity modulation, and spatial patterning of domains. The experiments are well designed and internally consistent. The switching is most effective near the CI/FCI fillings, is suppressed in gapless phases, and shows a power threshold and an electric-field window that match the gap-enhanced trion-lifetime picture. The photon-energy resonance with the trion is also checked. That is a convincing package; it is the first demonstration of optical writing of an FQAH-type ferromagnetic domain.\n\nThe main soft spot is the readout. The paper flips trion PL helicity and then states that the sign of the Chern number has switched. PL helicity is an established proxy for ferromagnetic polarization in this system, and ferromagnetic polarization is known to determine the Chern number sign, but the paper never puts a transport probe or any direct topological measurement on the optically switched state. The stress-test note is right: a helicity flip could in principle come from a local domain that does not propagate through the bulk, or from a metastable valley-polarized hole population under the beam spot. The inserted Fig. 3 phase-space comparison even shows that ρ behaves similarly in the CFL region near ν=-1/2, so a near-unity ρ after pumping does not by itself certify a Chern insulator. That weakens the title's claim of \"optical control of Chern insulators\" and turns it partly into an inference. It is a limitation, not a fatal flaw, because prior work has established the PL-helicity-to-Chern-number link in this material, but the language should be softened or the claim backed by a direct measurement.\n\nMinor technical point: the manuscript contains two captions labeled Fig. 3, with the second one appearing to be an inserted phase-space comparison. That is a copyediting problem.\n\nOverall: the core observation is novel, the data are careful, and the mechanism is credible. The paper deserves a serious referee. A good referee should ask for either a transport measurement of the switched state or a clear statement that the topological claim rests on the PL proxy. I would cite this work if I were working in moiré quantum matter or optical control of topological phases.\n\nRecommendation: send to peer review, and require the authors to address the readout question head-on.","headline":"A strong experimental demonstration of optical writing and erasing of moiré ferromagnetism at integer and fractional Chern insulator fillings, with the topological conclusion riding on the established PL-helicity proxy rather than a direct transport readout.","tokens_in":13114,"tokens_out":2022,"would_cite":true,"duration_ms":20183,"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":"The paper demonstrates that circularly polarized light tuned to the trion resonance flips the ferromagnetic polarization of twisted bilayer MoTe$_2$ and thereby reverses the sign of the Chern number of both its integer ($\\nu=-1$) and…","keywords":["twisted bilayer MoTe2","fractional Chern insulator","quantum anomalous Hall effect","optical switching","trion resonance","moiré ferromagnetism","circular dichroism","Chern number control"],"falsifier":"Apply the optical pumping protocol and then measure the Hall resistance at the same filling without changing anything else: if the $\\nu=-1$ state does not show a quantized anomalous Hall plateau of the opposite sign, and the $\\nu=-2/3$ state does not show the corresponding opposite fractionally quantized Hall resistance, then switching of the Chern number is not established even if the photoluminescence helicity flips.","tokens_in":12101,"feed_emoji":"🔄","tokens_out":14339,"duration_ms":122413,"temperature":0.7,"pith_summary":"Circularly polarized light tuned to the trion resonance can deterministically flip the ferromagnetic polarization of twisted bilayer MoTe$_2$, and with it the sign of the Chern number of both the integer Chern insulator at $\\nu=-1$ and the fractional Chern insulator at $\\nu=-2/3$. The paper demonstrates two protocols: optical training, in which the light is applied while the density is swept into the topological state, and direct switching at a fixed filling. Both produce complete reversal of the degree of circular polarization of the trion photoluminescence; toggling the pump helicity switches the state back and forth, and focused optical spots write stable domains of opposite Chern number. If correct, this is a non-thermal, zero-magnetic-field optical method for controlling and patterning strongly correlated topological phases.","feed_headline":"Polarized light flips Chern number in twisted MoTe2","feed_subtitle":"Trion-resonant light flips the magnetization of twisted MoTe2, enabling optical writing of integer and fractional Chern insulator domains.","key_machinery":"The machinery is the trion resonance, the optical transition in which a photoexcited electron-hole pair binds to a hole already present in the valence band. A $\\sigma^-$ pump creates valley-polarized electron-hole pairs in the $-K$ valley; the electron scatters intervalley more readily than the hole, so the time-reversed trion forms and recombines, depleting the hole population in the $+K$ valley. The net transfer of holes between valleys acts as an optical torque that can overcome the magnetic anisotropy barrier. The paper argues that an energy gap at the Chern insulator fillings lengthens the trion lifetime, increasing the optical torque and explaining why training and switching are most effective at $\\nu=-1$ and $\\nu=-2/3$.","core_discovery":"The paper's central claim is that circularly polarized light resonant with the trion transition acts on twisted bilayer MoTe$_2$ as an optical torque on its moiré Chern ferromagnet, and that applying that torque reverses the ferromagnetic polarization and hence flips the sign of the Chern number of the topological states it supports. In the authors' words, the complete switching of trion photoluminescence helicity demonstrates “the flipping of the ferromagnetic polarization, and thus the sign of the underlying Chern numbers for both CI and FCI states.” The effect is shown in two regimes: optical training, where the pump is on while the density is swept from a non-ferromagnetic value into the $\\nu=-1$ or $\\nu=-2/3$ state, and direct switching at fixed filling. Switching is most effective at the gapped Chern states, is reversible by toggling the pump helicity, and can be done locally to write ferromagnetic and Chern domains that persist for many hours.","pith_inferences":["The Chern-number conclusion inherits the assumption that PL helicity is a one-to-one readout of ferromagnetic polarization; a transport measurement of the switched state would directly confirm the sign change of the quantized Hall conductance, a check the paper does not perform.","If the mechanism is valley-selective hole injection, the same optical torque might also work at other gapped correlated fillings in the moiré flat band, including fractional states beyond $\\nu=-2/3$, although the paper only demonstrates $\\nu=-1$ and $\\nu=-2/3$.","Writing adjacent domains with Chern numbers of opposite sign creates chiral interfaces; for the $\\pm 2/3$ FCI case those interfaces could in principle support fractionally charged chiral edge modes, opening a route to light-defined fractional edge-state devices.","The power threshold depends on trion lifetime, so engineering the optical environment (for example, cavity or waveguide coupling) is a natural, untested route to lower the switching power."],"forward_implications":["Optical training gives deterministic on-demand initialization of QAH and FQAH polarization at zero magnetic field, replacing the magnetic-field poling step.","Alternating the pump helicity toggles the ferromagnetic polarization and therefore the Chern number, demonstrating an optically rewritable topological memory.","Focused pumping writes opposite-Chern-number domains that are stable for at least twenty hours; if domain walls carry chiral edge currents, this patterns edge-state channels.","Since about 1 eV photons control meV-scale many-body gaps at 1.6 K, the switching mechanism is non-thermal and does not rely on heating near the Curie temperature.","Devices made from higher-quality crystals switch at lower power, indicating that defect-controlled spin/valley relaxation is a practical tuning knob for optical control."],"supporting_citations":[{"why":"Prior observation of the fractionally quantized anomalous Hall effect in twisted bilayer MoTe$_2$; supplies the transport identification of the FCI states and the electric-field dependence of the Chern gap.","marker":"[15]"},{"why":"Device 1 is the same twisted MoTe$_2$ device used here; supplies the QAH and FQAH signatures at $\\nu=-1$ and $-2/3$ that define the states being optically controlled.","marker":"[17]"},{"why":"Establishes the gate-tunable correlated magnetic states and excitonic response of the material that make helicity-selective optical access possible.","marker":"[29]"},{"why":"Establishes that trion photoluminescence helicity is determined by ferromagnetic polarization, the calibration that converts PL polarization into a Chern-number readout.","marker":"[31]"},{"why":"Prior demonstration of coupling between light helicity and the QAH state in moiré Chern magnets; the precedent for light-controlled Chern magnetism.","marker":"[33]"},{"why":"High-quality crystal growth used for Device 2; its low defect density is what lowers the optical switching power threshold.","marker":"[34]"},{"why":"Demonstrates quantized chiral edge conduction on domain walls of a magnetic topological insulator; the basis for expecting edge currents at optically written domain boundaries.","marker":"[32]"}],"fun_headline_variants":["Light helicity flips Chern number in twisted MoTe2","Twisted MoTe2 Chern states flipped by light","Optical switch for fractional Chern insulator","Light rewrites Chern number domains in tMoTe2","Helicity-controlled Chern switch in twisted MoTe2"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the handedness of the emitted trion light is a reliable, one-to-one report of which way the ferromagnetic polarization points, and therefore of the sign of the topological invariant; the paper reads out magnetization only through photoluminescence helicity and never verifies the switched state by transport.","fun_headline_variants_meta":{"raw":{"variants":["Light helicity flips Chern number in twisted MoTe2","Twisted MoTe2 Chern states flipped by light","Optical switch for fractional Chern insulator","Light rewrites Chern number domains in tMoTe2","Helicity-controlled Chern switch in twisted MoTe2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000867,"raw_usage":{"total_tokens":3808,"prompt_tokens":1044,"completion_tokens":2764,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":660,"completion_tokens_details":{"reasoning_tokens":2688}},"tokens_in":660,"tokens_out":2764,"duration_ms":19158,"temperature":1.0,"reasoning_tokens":2688,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:53:53.940703+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Apply the optical pumping protocol and then measure the Hall resistance at the same filling without changing anything else: if the $\\nu=-1$ state does not show a quantized anomalous Hall plateau of the opposite sign, and the $\\nu=-2/3$ state does not show the corresponding opposite fractionally quantized Hall resistance, then switching of the Chern number is not established even if the photoluminescence helicity flips.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior observation of the fractionally quantized anomalous Hall effect in twisted bilayer MoTe$_2$; supplies the transport identification of the FCI states and the electric-field dependence of the Chern gap."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the gate-tunable correlated magnetic states and excitonic response of the material that make helicity-selective optical access possible."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Demonstrates quantized chiral edge conduction on domain walls of a magnetic topological insulator; the basis for expecting edge currents at optically written domain boundaries."}],"review_version":2}