{"id":"fc5411ed-8671-456a-bf8d-828400343e34","arxiv_id":"2507.17146","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Photo-doping Ca2RuO4 creates a metastable state at 10 microseconds with broken glide-plane symmetry and a magnetic response consistent with intra-layer ferromagnetic order.","lead":"Shining ultrafast laser pulses on the Mott insulator Ca2RuO4 creates a magnetic state that appears only in an intermediate time window, around 10 microseconds after excitation. The result points to a neglected timescale where photo-excited materials can hide new phases.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 10 µs 'intermediate time' claim rests on a stroboscopic pre-pump snapshot with acknowledged multi-pulse persistence; no direct time-resolved data establish that the FM-like glide-broken state emerges in the 100 ps–10 µs window after a single pulse.","rationale":"The reader's weakest assumption is substantially correct: the stroboscopic 10 µs point is taken in a regime where the sample does not fully recover between pulses, and the paper itself acknowledges this in Section S12. I treat this as the load-bearing concern because the title and abstract make 'intermediate timescales' central to the claimed novelty. If the state is instead an accumulated multi-pulse steady state, or if it actually forms on sub-nanosecond timescales and merely persists, the temporal narrative is weakened even though the existence of a metastable, glide-broken magnetic state would not be disproven. The reader also flags the borrowed Sun–Millis trapping mechanism; I regard that as a secondary interpretative issue because the empirical state is established by symmetry-resolved SHG and birefringence independent of the mechanism. The paper deserves credit for careful symmetry analyses, multidomain checks, CW-pump heating controls, and the fluence-temperature phase boundary mapping, all of which support the existence of a genuinely new metastable state. The missing piece is direct time-resolved evidence for when that state emerges after a single excitation, and the proposed low-repetition-rate or single-shot delay scan would settle it. Therefore the conditional verdict remains appropriate, with no adjustment beyond what the reader already recommended.","tokens_in":33199,"tokens_out":11868,"duration_ms":137051,"concrete_test":"Acquire time-resolved ηFM (BFISH or differential birefringence) after a single pump pulse, or at a repetition rate low enough (about 1 kHz or below) with probe delays scanned across 100 ps to 10 µs, for example via electronic delay or asynchronous sampling. If ηFM remains zero until microsecond delays and then grows, the intermediate-timescale emergence is confirmed. If ηFM is already maximal at about 1 ns after one pulse, or appears only after many pulses, the 'time-hidden intermediate window' claim and the per-cycle trapping scenario in Fig. 4e are not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that photo-doping creates a state 'emerging at intermediate timescales' depends on assigning the 10 µs BFISH/birefringence signal to the state reached about 10 µs after a single excitation pulse. Section S12 concedes that at F > Fc the sample does not fully recover between 100 kHz pulses and that the metastable state is already present before the pump arrives. The 10 µs snapshot is therefore a pre-pump steady-state measurement in a pulse train, not a direct readout of single-pulse dynamics; the state could be built up over many pulses. Moreover, no data are shown between 100 ps and 10 µs: the 'slow domain coalescence' in Fig. 4e is inferred, not measured. If ηFM is already saturated at 1 ns after a single pulse, or if it requires many pulses to accumulate, the abstract's claim of a hidden state emerging in the previously unexplored intermediate window loses its temporal meaning. The existence of a metastable, glide-symmetry-broken magnetic state at 10 µs is well supported by the symmetry fits and birefringence; the insecure part is the time at which it emerges after a single excitation. The Sun–Millis trapping mechanism is likewise borrowed (Ref. 22) and not simulated for Ca2RuO4, but it is an interpretative overlay rather than the core empirical finding.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports time-resolved second-harmonic generation (SHG) and optical birefringence experiments on the multi-orbital Mott insulator Ca2RuO4. After intense photoexcitation above a critical fluence, the authors observe at a nominal 10 μs delay a state in which the bc glide-plane symmetry is broken, as evidenced by rotational-anisotropy SHG patterns that require a surface magnetic point group 1 tensor fit, and by a bulk-sensitive birefringence asymmetry that is zero in equilibrium and grows above a fluence threshold. The magnetic-field-induced SHG data are interpreted as consistent with intra-layer ferromagnetic order with moments along the c-axis, arising after the intra-layer antiferromagnetic order has melted. A two-site exact-diagonalization model shows that reducing the orbital splitting Δ drives a ground-state crossover from intra-layer AFM to intra-layer FM correlations, and a mean-field calculation shows that photodoping reduces Δ. The authors conclude that photo-doping accesses a metastable 'time-hidden' magnetic state on the intermediate microsecond timescale, distinct from all equilibrium phases.","tokens_in":33571,"tokens_out":7054,"duration_ms":74877,"significance":"If the identification holds, this is a striking result: a metastable, glide-symmetry-broken magnetic state in a Mott insulator that appears on a microsecond timescale, populating a previously little-explored time window between ultrafast and quasi-static regimes. The paper's strengths are its internal consistency and cross-validation: the equilibrium AFM phase is fit with the expected magnetic point group m, the metastable phase requires point group 1, the same conclusion is reached by surface-sensitive SHG and bulk-sensitive birefringence, and the domain-reversal measurements support an ordered state rather than a heating artifact. The two-site model is solved by exact diagonalization and inherits parameters from prior work, with only one fitted spring constant in the mean-field analysis. The main weakness is that the temporal claim of emergence in the intermediate window is not directly established by the data, as discussed below; the state is well characterized at 10 μs under repetitive excitation, but its single-pulse evolution from 100 ps to 10 μs is inferred rather than measured.","major_comments":[{"comment":"The central 'intermediate time window' claim is not directly established by the data. The 10 μs data (e.g., Fig. 2d(iii)) are acquired with the probe pulse arriving slightly before the pump in a 100 kHz train, and Section S12 concedes that for F > Fc the sample does not fully recover between pulses and that the metastable state is already present before the pump arrives. The measurement therefore reports a steady-state, multi-pulse condition rather than the state reached about 10 μs after a single excitation pulse; the signal could accumulate over many cycles. No data are shown between 100 ps and 10 μs, so the slow domain-coalescence evolution in Fig. 4e is inferred, not measured. The abstract's claim that the state 'emerges' in the previously unexplored intermediate window should either be supported by single-pulse or variable-repetition-rate experiments, or be rephrased to describe a metastable state sustained under repetitive excitation.","section":"Fig. 4e and Section S12"},{"comment":"The trapping mechanism is invoked but not demonstrated for Ca2RuO4. The two-site exact-diagonalization calculation establishes that the FM state is a local minimum for Δ below Δc, and the Section S2 mean-field calculation shows that photodoping reduces Δ, but neither calculation simulates the relaxation dynamics or the population of the FM valley after the potential is restored. The statement in Fig. 4e that the system is trapped 'possibly through the transient trapping mechanism proposed by Sun and Millis [22]' is an untested assumption, and the word 'trajectory' in the abstract is stronger than what the calculations show. The authors should either perform a time-dependent simulation of the relaxation or explicitly label this step as a hypothesis.","section":"Fig. 4e and Section S9"},{"comment":"The magnetic nature of the new state is inferred from symmetry fits rather than measured by a direct magnetic probe. The BFISH data are fit by a magnetic point group 1 tensor and the birefringence asymmetry shows glide-plane breaking, but no direct measurement of a c-axis ordered moment is presented; the absence of Kerr rotation in Section S7 is attributed to antiparallel stacking of FM planes. The abstract and main text are appropriately cautious in saying the observations are 'consistent with' intra-layer FM order, but the title 'Time-hidden magnetic order' goes beyond the directly measured quantities. A direct probe (e.g., time-resolved resonant x-ray scattering or a local magnetic probe) or a softening of the title and framing is needed.","section":"Sections S6, S7, and title"}],"minor_comments":[{"comment":"The citation of '[2, 12]' for the steep increase in Δ below TOO and for the 'previous neutron and x-ray diffraction measurements' is incorrect: reference [12] is Dean et al. on Sr2IrO4, not a Ca2RuO4 orbital-order study. Please replace with the appropriate Ca2RuO4 structural references.","section":"First paragraph and references"},{"comment":"The statement that the AFM-to-FM crossover as Δ is reduced is 'consistent with prior work [17]' appears to cite the wrong reference: [17] is a Keldysh control paper by Li et al., whereas the relevant prior two-site model is Meetei et al. (reference [8] in the main text or [16] in the Supplemental Information). Please correct this citation.","section":"Fig. 4c context, reference [17]"},{"comment":"The notation for the glide-symmetry-breaking order parameter is inconsistent: the main text uses ηFM (Fig. 2f) while Section S8 introduces ηM. Please unify the notation.","section":"Section S8 and main text"},{"comment":"The phrase '10 μs snapshot was obtained by having the probe pulse arrive slightly before the pump which, in a repetitive measurement, is equivalent to setting the time delay to the laser pulsing period' should explicitly note that this only samples the pre-pump steady state if the system has reached a periodic steady state; the associated assumption is acknowledged in S12 but should also be flagged in the main text.","section":"Methods, time-resolved SHG"}],"recommendation":"major_revision","confidential_remarks":"The experimental core of the paper is convincing and the symmetry analysis is careful. The main revision issue is the temporal interpretation: the 10 μs state is well characterized but the single-pulse emergence claim is not supported by the current stroboscopic scheme. The authors should be asked to provide single-pulse or variable-repetition-rate evidence, or to explicitly reframe the central claim as a state observed under repetitive excitation. The magnetic assignment would also benefit from a direct probe or from more cautious wording in the title."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the core experimental result is genuinely new: a metastable, glide-symmetry-broken magnetic state appears in photo-doped Ca2RuO4 under 100 kHz pumping and is consistent with intra-layer FM order. The symmetry analysis is careful and cross-validated between surface SHG and bulk birefringence, including domain reversal. That state is not in the equilibrium phase diagram, and prior theory (including the paper's own two-site model, which credits Meetei et al.) predicted a Delta-tuned AFM-FM transition but not this photo-induced metastable realization. So credit where due: the experiment is well executed and the identification is about as good as optics can do without direct magnetization measurement.\n\nNow the soft spot, and it's real. The paper claims the state emerges in the previously unexplored intermediate window (roughly 100 ps to 10 µs). But the 10 µs data point is a stroboscopic pre-pump snapshot in a 100 kHz pulse train. Section S12 concedes the sample does not fully recover between pulses and that the metastable state is already present before the pump arrives. So the 10 µs measurement is a steady-state property of the pumped sample, not a single-pulse dynamic. There is no data between 100 ps and 10 µs; the 'slow domain coalescence' in Fig. 4e is a hypothesis. If the state needs several pulses to accumulate, or if it is already saturated at 1 ns, the 'intermediate time' framing loses its content. This is a load-bearing narrative issue, not a fatal blow to the existence of the state.\n\nMinor points: the FM assignment is inferred from the departure of BFISH patterns from surface group m, which is consistent but not a direct probe of magnetization; and the Sun-Millis trapping mechanism is borrowed without simulation for this material. Both are acknowledged in the text.\n\nBottom line: the paper deserves a serious referee. The experimental phase is solid enough for publication, but the temporal claim needs rework — either measure the 0.1–10 µs window directly, or reframe the result as a persistent metastable state under repetitive pulsing. I'd send it to review with a referee who will press on that distinction.","headline":"Real metastable FM-like state found at 10 µs, but the 'intermediate timescale' emergence claim rests on a stroboscopic pre-pump snapshot, not on resolved dynamics.","tokens_in":34105,"tokens_out":2307,"would_cite":true,"duration_ms":26225,"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":"Photo-doping Ca2RuO4 produces a metastable intra-layer ferromagnetic state at 10 microseconds, after antiferromagnetic order has melted and carriers have recombined, with broken glide-plane symmetry.","keywords":["Ca2RuO4","Mott insulator","photo-doping","metastable magnetic order","second-harmonic generation","ferromagnetism","intermediate timescale","orbital splitting"],"falsifier":"Run the same pump-probe experiment at a repetition rate low enough (for example 1 kHz) that the sample fully relaxes between pulses, and look for the ferromagnetic signal 10 microseconds after a single pump pulse; if the signal disappears, the 10 microsecond state is a multi-pulse artifact rather than the single-cycle intermediate-time state.","tokens_in":33050,"feed_emoji":"🧲","tokens_out":8052,"duration_ms":82508,"temperature":0.7,"pith_summary":"This paper reports that shining near-infrared pulses on the Mott insulator Ca2RuO4 creates a magnetic state that has no counterpart in the equilibrium phase diagram. The state is detected 10 microseconds after excitation, a time window usually skipped between ultrafast and long-lived regimes. At that moment the usual antiferromagnetic order has already melted and the photo-created carriers have already recombined, and the optical signatures point to intra-layer ferromagnetic order with the crystal's glide plane broken. The authors propose a non-thermal route: photo-doping temporarily shrinks the orbital splitting below a critical value, trapping the system in a ferromagnetic valley of its energy landscape. If right, this expands the search for hidden electronic phases to intermediate timescales where states can be transiently trapped.","feed_headline":"Photo-doping makes Ca2RuO4 a hidden ferromagnet at 10 microseconds","feed_subtitle":"Appears only after antiferromagnetic order melts and photocarriers recombine, and is thermally inaccessible.","key_machinery":"The control parameter is the orbital splitting between the dxy orbital and the dyz/xz doublet in the flattened RuO6 octahedra. Photodoping reduces that splitting by lowering the charge gap, and when it drops below a critical value the intra-layer magnetic ground state switches from antiferromagnetic (moments along b) to ferromagnetic (moments along c). The experiments use two bulk-sensitive probes: magnetic-field-induced second-harmonic generation, whose time-reversal-odd tensor records the magnetic point group, and modulation-based differential birefringence, which detects loss of the bc glide plane. The timescale argument is carried by the transient-trapping mechanism proposed for systems with competing orders, in which the system is captured in a local valley while the original potential landscape is restored. The two-site exact-diagonalization model supplies the splitting-dependent energy landscape and the weak azimuthal anisotropy that explains slow domain growth.","core_discovery":"The central claim is that photo-doping Ca2RuO4 at fluences above a threshold melts the intralayer antiferromagnetic order within about 2 ps and then leaves the crystal in a metastable state that, by 10 microseconds, shows second-harmonic and birefringence signatures incompatible with any equilibrium phase: the bc glide plane is broken and the magnetic point group drops from m to 1, consistent with intra-layer ferromagnetic moments pointing along the c-axis. The state appears only after the antiferromagnetic order parameter has collapsed and after photocarriers have recombined, and only below the Neel temperature; it is insulating, and it relaxes back to the antiferromagnetic state within a minute. Exact-diagonalization of a two-site model of the RuO6 network shows that reducing the orbital splitting below a critical value switches the ground state from antiferromagnetic (b-axis) to ferromagnetic (c-axis), and that the ferromagnetic valley has very weak in-plane anisotropy, so small domains can take much longer than electronic relaxation times to grow into a detectable net signal.","pith_inferences":["The paper's 10 microsecond point is a stroboscopic pre-pump snapshot in a 100 kHz pulse train; the claim that this is the single-pulse intermediate-time state would be directly tested by a low-repetition-rate or single-shot measurement that lets the sample fully relax between pulses.","If the splitting-controlled antiferromagnetic-to-ferromagnetic switch is generic, other d4 or multi-orbital Mott insulators with a similar orbital splitting might host analogous time-hidden ferromagnetic states, and epitaxial strain or chemical pressure might stabilize them in equilibrium.","Because optical measurements determine symmetries but not atomic positions, a neutron or resonant X-ray scattering experiment timed to the 10 microsecond window would be a cleaner confirmation of the c-axis ferromagnetic moment arrangement.","The weak in-plane anisotropy predicted by the two-site model implies the hidden order should appear as small domains first; spatially resolved probes could observe domain growth directly rather than infer it from the integrated signal."],"forward_implications":["Microsecond time delays become a valid hunting ground: hidden states in driven quantum materials can appear after electronic and phononic relaxation has finished, when domain coarsening or trapping has had time to act.","A fluence threshold and a temperature below the Neel point are required: the intra-layer ferromagnetic state appears only once the pump has melted the antiferromagnetic order and only if the system starts from the antiferromagnetic phase, not from the paramagnet.","The new state is thermally inaccessible in equilibrium and is insulating, so it is a different object from the pressure- or strain-driven itinerant ferromagnetism previously reported in Ca2RuO4.","The observable onset of the hidden order is set by domain growth, not by the initial electronic relaxation, so optical signals can stay zero for nanoseconds and then rise on the microsecond scale.","Within one pulse cycle the state forms, persists for an appreciable fraction of the 10 microsecond period, and then the crystal recovers to the antiferromagnetic state within about a minute, defining the metastable lifetime."],"supporting_citations":[{"why":"Predicts that reducing the structural distortion switches Ca2RuO4 from intra-layer antiferromagnetic to intra-layer ferromagnetic order, supplying the theoretical foundation for the metastable state.","marker":"[7]"},{"why":"Establishes the orbital ordering transition and the steep rise of the orbital splitting below the ordering temperature, which lets the authors track the splitting through second-harmonic intensity.","marker":"[2]"},{"why":"Shows ultrafast melting of magnetic correlations in a photodoped Mott insulator, supporting the observed 2 ps collapse of antiferromagnetic order.","marker":"[12]"},{"why":"Supplies the transient-trapping mechanism invoked to explain how the system relaxes into the ferromagnetic valley rather than returning to the antiferromagnetic ground state.","marker":"[22]"},{"why":"Provides the Hund's-coupling framework that favors ferromagnetic correlations in multi-orbital systems away from half-filling, underpinning the model intuition.","marker":"[20]"},{"why":"Supplies pump-probe optical-conductivity data showing the crystal remains Mott insulating at fluences above the critical value, ruling out an insulator-metal transition.","marker":"[17]"},{"why":"Documents recombination and the decoupling of static and dynamic criticality in a driven Mott insulator, supporting the claim that carriers have recombined before 10 microseconds.","marker":"[11]"}],"fun_headline_variants":["Photo-doped Ca2RuO4 hides a new ferromagnetic state","Ca2RuO4 reveals hidden ferromagnetism after photo-doping","Hidden ferromagnet emerges in Ca2RuO4 at 10 μs","Light exposure turns Ca2RuO4 into a hidden ferromagnet","Photo-doping uncovers a fleeting ferromagnetic phase in Ca2RuO4"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The 10 microsecond snapshot is taken while a 100 kHz pulse train is still hitting the sample, and the paper's Section S12 concedes that the crystal does not fully recover between pulses; the whole picture assumes that what is seen at 10 microseconds is the state produced within one excitation cycle rather than an accumulated multi-pulse effect.","fun_headline_variants_meta":{"raw":{"variants":["Photo-doped Ca2RuO4 hides a new ferromagnetic state","Ca2RuO4 reveals hidden ferromagnetism after photo-doping","Hidden ferromagnet emerges in Ca2RuO4 at 10 μs","Light exposure turns Ca2RuO4 into a hidden ferromagnet","Photo-doping uncovers a fleeting ferromagnetic phase in Ca2RuO4"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000945,"raw_usage":{"total_tokens":4044,"prompt_tokens":961,"completion_tokens":3083,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":577,"completion_tokens_details":{"reasoning_tokens":2984}},"tokens_in":577,"tokens_out":3083,"duration_ms":23213,"temperature":1.0,"reasoning_tokens":2984,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T14:56:52.144932+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the same pump-probe experiment at a repetition rate low enough (for example 1 kHz) that the sample fully relaxes between pulses, and look for the ferromagnetic signal 10 microseconds after a single pump pulse; if the signal disappears, the 10 microsecond state is a multi-pulse artifact rather than the single-cycle intermediate-time state.","supporting_citations":[{"cited_title":"& Terakura, K","cited_arxiv_id":null,"evidence_quote":"Predicts that reducing the structural distortion switches Ca2RuO4 from intra-layer antiferromagnetic to intra-layer ferromagnetic order, supplying the theoretical foundation for the metastable state."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the orbital ordering transition and the steep rise of the orbital splitting below the ordering temperature, which lets the authors track the splitting through second-harmonic intensity."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows ultrafast melting of magnetic correlations in a photodoped Mott insulator, supporting the observed 2 ps collapse of antiferromagnetic order."},{"cited_title":"& Millis, A","cited_arxiv_id":null,"evidence_quote":"Supplies the transient-trapping mechanism invoked to explain how the system relaxes into the ferromagnetic valley rather than returning to the antiferromagnetic ground state."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Hund's-coupling framework that favors ferromagnetic correlations in multi-orbital systems away from half-filling, underpinning the model intuition."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents recombination and the decoupling of static and dynamic criticality in a driven Mott insulator, supporting the claim that carriers have recombined before 10 microseconds."}],"review_version":1}