{"id":"5aef35a7-9509-4256-bc15-d839cce19e98","arxiv_id":"1908.06413","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In EuNiO3/LaNiO3 superlattices, suppressing long-range charge ordering reduces low-frequency resistance noise by about three orders of magnitude in the insulating phase, while phase-separation physics near the transition remains similar.","lead":"This paper measures electrical noise in two types of nickelate superlattices as they switch between metal and insulator. It finds that the sample without long-range charge ordering shows about a thousand times less noise in the insulating state, suggesting a way to make cleaner electronic devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The 10^3 noise difference is attributed to charge ordering, but the two films differ in composition, interface density, and growth history, with no error bars or reproducibility check; a direct control is missing.","rationale":"The paper's qualitative observations—RTN, non-1/f spectra, and non-Gaussian second spectrum near TMIT—are internally consistent and support electronic phase separation in both systems. The strongest quantitative claim is the 10^3 noise reduction, and that claim is load-bearing for the device-relevance message. The missing control is not a matter of disagreement with consensus; it is a direct threat to soundness because 1/f noise is a sensitive probe of sample microstructure. The reader's weakest assumption identifies this correctly. I would not reject the paper; the conclusion is plausible and the methods are standard. However, the absence of error bars and sample-to-sample reproducibility means the headline number should be treated as preliminary. A secondary concern: the domain-size estimate in Section III, using Ea ~ 0.42 eV and Ev ~ 160-190 kJ/m^3 with a spherical-volume assumption, yields a diameter of about 9 nm rather than the quoted 7.0-7.4 nm when checked with the formula L = (6Ea/pi*Ev)^(1/3). This does not change the qualitative phase-separation picture but suggests the quantitative length scale is less robust than stated.","tokens_in":10935,"tokens_out":8046,"duration_ms":81478,"concrete_test":"Grow at least two additional 1ENO/1LNO and two additional 2ENO/1LNO superlattices under nominally identical PLD conditions and measure the integrated normalized noise in the insulating plateau (e.g. T/TMIT ~ 0.6) for all six samples, computing 95% confidence intervals from segment-to-segment variability. If feasible, also grow a control pair that keeps the cation ratio fixed while changing only the superlattice periodicity mismatch relevant to CO (e.g. 1ENO/2LNO vs 2ENO/1LNO). If the within-sample-type scatter is comparable to or larger than the claimed 10^3 gap, or if the control pair does not reproduce the gap, the attribution to long-range CO is not supported. As an immediate first check, re-analyze the existing time series using jackknife resampling to obtain error estimates on every quoted delta-R^2/R^2 point.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that suppressing long-range CO lowers insulating-phase noise by about 10^3 (Fig. 3 insets)—depends on the two superlattices being identical in every respect except for the presence of long-range CO. They are not shown to be: 2ENO/1LNO and 1ENO/1LNO differ in nominal Eu:La ratio (2:1 vs 1:1), superlattice periodicity (3 uc vs 2 uc), number of interfaces (24 vs 36), and transition temperatures (TMIT = 245 K vs 165 K). The absence of long-range CO is taken from Refs. 21, 39, and 40 rather than from characterization of the exact films used here. Low-frequency noise magnitude is highly sensitive to defect density, interface roughness, and disorder, so any of these uncontrolled differences could produce a large offset in normalized noise. The paper reports no error bars on the integrated noise values and no repeat measurements on an independently grown sample of either superlattice, so the statistical significance of the claimed 10^3 separation is unverified. The comparison is therefore an uncontrolled single-pair experiment; the conclusion may be correct, but the data as presented cannot exclude a growth- or interface-related origin for the difference.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper reports low-frequency resistance noise measurements on two [EuNiO3/LaNiO3] superlattices: 1ENO/1LNO, which exhibits a simultaneous metal-insulator and charge-ordering transition near 165 K, and 2ENO/1LNO, which undergoes a MIT near 245 K without long-range charge ordering. The authors observe random telegraphic noise, non-1/f power spectra, and non-Gaussian second spectra near the transitions in both samples, extract a common activation energy Ea ≈ 0.42 eV from the Lorentzian corner frequency, and estimate metallic domain diameters of about 7 nm. They also report that the integrated noise in the insulating phase is about three orders of magnitude lower in 2ENO/1LNO than in 1ENO/1LNO, and they suggest that suppressing long-range charge order reduces excess noise and may be beneficial for device applications.","tokens_in":11177,"tokens_out":5334,"duration_ms":55763,"significance":"If the central comparison were properly controlled, the paper would be significant: it would show that noise spectroscopy can distinguish the role of long-range charge order from the generic phase-separation dynamics in nickelate heterostructures, and it would support digital synthesis as a route to low-noise oxide electronics. The experimental methodology is standard for noise spectroscopy, and the concurrent appearance of RTN, non-1/f spectra, and non-Gaussian second spectra is internally consistent evidence for electronic phase separation in both systems. The similar Ea and inferred length scale in the two samples are also valuable observations. However, the main quantitative claim—the 10^3 reduction in insulating-phase noise—currently rests on a single pair of films with no statistical validation, so the significance is conditional on additional control measurements.","major_comments":[{"comment":"The claimed three-orders-of-magnitude difference in insulating-phase noise between 1ENO/1LNO and 2ENO/1LNO is not statistically supported: no error bars are given for the integrated noise values, no repeat measurements on independently grown films are reported, and the two superlattices differ in Eu:La ratio, periodicity, number of interfaces, and TMIT. Since low-frequency noise magnitude is highly sensitive to defect density and interface disorder, these data cannot exclude a growth- or microstructure-related origin for the offset; the authors' own sentence 'We speculate that because of the absence of long range CO in 2ENO/1LNO SL, noise magnitude is smaller than 1ENO/1LNO SL' correctly identifies this as an open point. This comparison is load-bearing for the device-oriented conclusion.","section":"Section III, Fig. 3(a),(b) insets and text beginning 'The PSD of resistance fluctuations ...'"},{"comment":"The Arrhenius analysis yielding Ea = 0.42 ± 0.03 eV and the statement that this value is the same for both samples are presented without error bars on the individual fc values or on the fitted amplitudes A and B, and no statistical test of equality is provided. The claim of similar energy barriers in the two samples is used to support the interpretation of common phase-separation physics, so this should be quantified.","section":"Section III, Eq. (1) and Fig. 2(d)"},{"comment":"The estimate Lm ≈ 7.0–7.4 nm assumes that Ea is purely elastic energy from the volume mismatch, uses bulk moduli of the two constituents and a 0.2% out-of-plane expansion, and assumes spherical metallic nuclei. These are strong assumptions; Ea could include electronic or magnetic contributions, and the inferred domain size is not independently verified. The sentence 'Our results emphasizes that nucleation of such metallic phase happens at much shorter length scale' should be softened or supported by additional evidence.","section":"Section III, paragraph beginning 'The length scale associated with the electronic phase separation ...'"}],"minor_comments":[{"comment":"The exponent w ≈ 2 ± 0.1 is quoted from a log-log fit of δR²/R² versus R, but the fitting range, number of points, and confidence intervals are not stated; please report these details.","section":"Appendix, Fig. 6 and the percolation paragraph"},{"comment":"The sentence 'At this moment, the reason for this shift between the transition temperature obtained from resistivity measurement and the temperature of noise peak remains unclear' leaves an unresolved interpretive point; if the additional noise peak near 210 K in 2ENO/1LNO is presented as a result, a brief discussion of possible systematic offsets would be helpful.","section":"Section III, Fig. 3 discussion"},{"comment":"The statement that σ(2) = 3 for Gaussian fluctuations is an ideal-limit result; the finite octave bandwidth and measurement noise floor can bias the baseline, so a short comment on the expected Gaussian value under the present measurement conditions would strengthen the non-Gaussian claim.","section":"Section III, second spectrum definition"},{"comment":"The Hooge parameter comparison across different nickelate systems should note that the values were obtained under different growth, geometry, and measurement conditions; a direct tabular comparison without this caveat may overstate the differences.","section":"Table I"}],"recommendation":"major_revision","confidential_remarks":"The paper is a technically solid noise-spectroscopy study, but the headline 10^3 comparison needs a control: either repeat measurements with error bars, a second pair of samples, or a direct structural characterization of the exact films used. I would support reconsidering a revised version in which the authors either supply that evidence or explicitly decouple the measured observation from the causal attribution to charge ordering."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Briefly: this is a clean noise-spectroscopy study of two EuNiO3/LaNiO3 superlattices, one with and one without long-range charge order, and the finding that the charge-order-free film has roughly 10^3 smaller low-frequency noise in the insulating phase is the kind of thing device people care about. The measurement itself is the new part: nobody has compared noise across a charge-order-suppressed nickelate superlattice before.\n\nWhat the paper does well: the within-sample analysis is careful and self-consistent. Both films show random telegraphic noise, a non-1/f Lorentzian component, and strongly non-Gaussian second spectra in the same reduced temperature window (0.85–1.1 T_MIT), with an activation energy around 0.42 eV and an inferred metallic-domain size around 7 nm. The percolation scaling (noise ~ R^w, w ~ 2) is consistent with the random-void model. The authors are appropriately careful to call the charge-order attribution speculation.\n\nThe soft spots are real but not fatal. The central 10^3 comparison rests on one pair of films. The two superlattices differ in Eu:La ratio, periodicity, interface density, and growth history, and the absence of long-range order is taken from prior structural work rather than measured on these exact films. There are no error bars on the integrated noise values and no repeat samples. Given how sensitive low-frequency noise is to defects and interfaces, an uncontrolled difference between the two growths could plausibly produce a large offset. So the claim 'CO suppression lowers noise' is plausible, but the data as presented do not demonstrate it beyond a single pair. The length-scale estimate also leans on assumed bulk moduli and a 0.2% lattice expansion; it is an order-of-magnitude estimate, not a measurement. The paper says 'we speculate,' which is honest, but a referee should ask for more.\n\nBottom line: this is a solid experimental contribution to the nickelate-noise literature and deserves a serious referee. The phase-separation picture for both films is well supported internally. The comparative claim needs either error analysis, a second growth, or a direct control before it can be taken as quantitative. I would send it out.","headline":"Useful comparative noise study with a plausible but unproven headline claim: the 10^3 noise reduction in the charge-order-free superlattice needs a statistical and sample-comparability check before it becomes quantitative.","tokens_in":11727,"tokens_out":2147,"would_cite":true,"duration_ms":22192,"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":"Suppressing long-range charge ordering in EuNiO3/LaNiO3 superlattices reduces insulating-phase resistance noise by three orders of magnitude, while electronic phase separation remains.","keywords":["resistance noise","1/f noise","metal-insulator transition","charge ordering","nickelate heterostructures","electronic phase separation","second spectrum","random telegraphic noise"],"falsifier":"A definitive test would be to grow a series of [mEuNiO3/nLaNiO3] superlattices with identical total volume and carrier density but differing periodicities, measure the insulating-phase noise, and confirm that it tracks the independently measured presence or absence of long-range charge ordering; if the noise contrast is found to depend on growth-induced disorder or defect density instead, the attribution would collapse.","tokens_in":10749,"feed_emoji":"⚡","tokens_out":9138,"duration_ms":76661,"temperature":0.7,"pith_summary":"This paper asks whether the charge-ordering transition that normally accompanies the metal-insulator transition in rare-earth nickelates is responsible for the enormous low-frequency resistance noise observed near the transition. By measuring resistance fluctuations in two engineered [EuNiO3/LaNiO3] superlattices—one with long-range charge ordering and one in which the superlattice periodicity suppresses it—the authors find that both systems exhibit random telegraphic noise, non-1/f spectra, and non-Gaussian second spectra, indicating electronic phase separation in both. The striking quantitative result is that the charge-order-free superlattice shows almost a thousand times smaller noise in its insulating phase, despite having a very similar activation energy (~0.42 eV) and metallic domain size (~7 nm). These findings matter because they suggest that digital synthesis can decouple the metal-insulator transition from charge ordering, yielding a correlated electron system with a sharp transition but much lower noise, which is preferable for device applications.","feed_headline":"Charge-order-free nickelates carry 1,000x less noise","feed_subtitle":"Noise spectroscopy shows both superlattices phase-separate, but only the one with long-range charge order is noisy.","key_machinery":"The central object is the low-frequency resistance-noise spectrum, measured with a four-probe lock-in technique as a time series of resistance fluctuations δR(t). From this time series the paper computes the power spectral density S_R(f), which near the transition decomposes into a 1/f component and a Lorentzian component with corner frequency f_c; the Arrhenius behavior of f_c gives the energy barrier E_a between metallic and insulating regions. The second spectrum, a four-point correlation function of the resistance fluctuations, is used to quantify non-Gaussian statistics; its normalized form σ(2) equals 3 for independent Gaussian fluctuators and deviates when correlated fluctuations are present. These tools are applied to a designed pair of superlattices that differ only in whether the superlattice periodicity matches the rock-salt charge-ordering periodicity, making the noise comparison attribution to charge ordering possible.","core_discovery":"The central discovery is that the presence or absence of long-range charge ordering changes the magnitude of resistance noise near the metal-insulator transition by about three orders of magnitude, while leaving the signatures of electronic phase separation essentially unchanged. In the charge-ordered 1ENO/1LNO superlattice, the integrated noise δR²/R² in the insulating phase is roughly 10³ times larger than in the 2ENO/1LNO superlattice that lacks long-range charge order. Both samples show random telegraphic noise over a comparable reduced-temperature window (T_RTN ~ 0.85 T_MIT), a power spectral density composed of a 1/f term and a Lorentzian with thermally activated corner frequency (E_a ~ 0.42 eV), and a normalized second spectrum σ(2) that deviates strongly from the Gaussian value of 3 near the transition. The noise magnitude scales as R² with exponent ~2, consistent with classical percolation of metallic clusters in an insulating matrix. The authors infer that the microscopic energetics of metallic cluster formation are similar in both cases, but the long-range charge-ordered state introduces additional, much stronger fluctuators.","pith_inferences":["An untested corollary of the similar activation energies is that the energy barrier for metallic cluster formation may be controlled by local epitaxial strain; varying the substrate to tune strain should then shift the noise peak and E_a, a measurement not performed in the paper.","Because the paper estimates metallic domains of only ~7 nm, far smaller than the ~100–300 nm domains seen by conductive atomic force microscopy in NdNiO3, noise spectroscopy may be uniquely sensitive to the earliest, nanometer-scale stages of phase separation; combining the two techniques on the same sample would test this.","The demonstrated low noise of the charge-order-free superlattice, if it holds under electrical cycling, suggests a path to using digital synthesis of nickelates in devices such as field-effect transistors or volatile switches, though the paper does not itself fabricate a device."],"forward_implications":["The 2ENO/1LNO superlattice combines a sharp first-order metal-insulator transition with roughly 10³ times lower insulating-phase noise than its charge-ordered counterpart, making it a better candidate for oxide electronic devices that switch through the MIT.","Observation of random telegraphic noise and non-Gaussian second spectrum in both samples implies that electronic phase separation is intrinsic to the nickelate MIT even when long-range charge ordering is absent.","The nearly identical activation energy (~0.42 eV) and domain size (~7 nm) in the two samples suggest that the local energetics of metallic cluster nucleation are set by short-range interactions rather than by the long-range charge-ordered state.","The noise peak and maximum of σ(2) near the Néel temperature in the 2ENO/1LNO sample indicate that the E'-type antiferromagnetic ordering also contributes to resistance fluctuations, possibly via Fermi surface nesting."],"supporting_citations":[{"why":"Establishes the two superlattice structures: 1ENO/1LNO has long-range charge ordering while 2ENO/1LNO does not, defining the comparison at the heart of the paper.","marker":"21"},{"why":"Supplies the lock-in based noise measurement and spectral analysis method used to obtain resistance fluctuation time series.","marker":"34"},{"why":"Provides the earlier observation of similar activation energy and large noise at charge-ordering transitions in manganites, the direct comparator for interpretation.","marker":"38"},{"why":"Defines the second spectrum formalism used to quantify non-Gaussian fluctuations.","marker":"33"},{"why":"Provides the random void model predicting noise scaling as R^w with w ~ 2.1, used to identify percolative transport.","marker":"46"},{"why":"Hall effect measurements showing a partially gapped Fermi surface, used to interpret the second-spectrum peak near the antiferromagnetic transition.","marker":"42"},{"why":"Supplies the bulk moduli of EuNiO3 and LaNiO3 used to convert the activation energy into the ~7 nm metallic domain size.","marker":"55"}],"fun_headline_variants":["No charge order, 1000x less noise in nickelates","Charge order boosts resistance noise 1000x","Nickelate without charge order: 1000x quieter","Charge-order-free nickelates: 1000x less noise","Eliminating charge order cuts noise 1000x"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The three-orders-of-magnitude noise difference is attributed solely to the presence versus absence of long-range charge ordering, which requires that the two films have identical volume, disorder, defect density, and measurement geometry.","fun_headline_variants_meta":{"raw":{"variants":["No charge order, 1000x less noise in nickelates","Charge order boosts resistance noise 1000x","Nickelate without charge order: 1000x quieter","Charge-order-free nickelates: 1000x less noise","Eliminating charge order cuts noise 1000x"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000526,"raw_usage":{"total_tokens":2553,"prompt_tokens":975,"completion_tokens":1578,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":591,"completion_tokens_details":{"reasoning_tokens":1495}},"tokens_in":591,"tokens_out":1578,"duration_ms":10947,"temperature":1.0,"reasoning_tokens":1495,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:45:58.683422+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A definitive test would be to grow a series of [mEuNiO3/nLaNiO3] superlattices with identical total volume and carrier density but differing periodicities, measure the insulating-phase noise, and confirm that it tracks the independently measured presence or absence of long-range charge ordering; if the noise contrast is found to depend on growth-induced disorder or defect density instead, the attribution would collapse.","supporting_citations":[{"cited_title":"Middey , author D","cited_arxiv_id":null,"evidence_quote":"Establishes the two superlattice structures: 1ENO/1LNO has long-range charge ordering while 2ENO/1LNO does not, defining the comparison at the heart of the paper."},{"cited_title":"Bid , author A","cited_arxiv_id":null,"evidence_quote":"Provides the earlier observation of similar activation energy and large noise at charge-ordering transitions in manganites, the direct comparator for interpretation."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the second spectrum formalism used to quantify non-Gaussian fluctuations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the random void model predicting noise scaling as R^w with w ~ 2.1, used to identify percolative transport."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Hall effect measurements showing a partially gapped Fermi surface, used to interpret the second-spectrum peak near the antiferromagnetic transition."},{"cited_title":"Eguchi , author A","cited_arxiv_id":null,"evidence_quote":"Supplies the bulk moduli of EuNiO3 and LaNiO3 used to convert the activation energy into the ~7 nm metallic domain size."}],"review_version":1}