{"id":"dc1cf8fe-086c-4ed4-9f6d-d4e421d355f2","arxiv_id":"2608.09841","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Dark-field X-ray microscopy maps the elastic strain and lattice rotation around individual threading dislocations in HgCdTe/CdZnTe epilayers, revealing strain lobes of about 5e-5 and a 7-micron misorientation signature equal to the layer thickness.","lead":"Scientists used a synchrotron X-ray microscope to map the tiny crystal distortions around individual defects in the infrared detector material HgCdTe. The maps show strain and rotation patterns that could help explain why some detector pixels have higher dark current.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed axial-strain lobes of ±(4–5)×10⁻⁵ sit within the reported matrix background of ±0.5×10⁻⁴, and no noise estimate is provided, so the central strain-field result is not yet distinguishable from background.","rationale":"I read the paper in good faith. The DFXM experiment is well conceived, the reflection-geometry setup is plausible, and the multiscale KAM images of individual threading dislocations are a useful advance for MCT/CZT. The internal Bragg-angle inconsistency identified by the reader is real but not load-bearing: the geometric conclusion that a vertical line of length 7 µm projects to a detector-plane separation of t cosθ_B is essentially unchanged for θ_B=5.605° versus θ_B≈8.8°, since both cosines are within 1% of unity. The more serious issue is that the reported strain-lobe amplitudes are comparable to the reported background variation of the same ε33 map, and no uncertainty analysis is provided. Because the paper's central novelty is the separation of elastic strain from lattice rotation, a strain signal that could be from the acknowledged substrate-convolution background or from first-moment noise would weaken the claim substantially. The reader's verdict of CONDITIONAL is appropriate; my concern reinforces the need for a quantitative noise characterization before the strain-field result is treated as authoritative. I therefore do not change the verdict, but I flag a different weakest point than the reader's projection-formula concern.","tokens_in":8682,"tokens_out":13089,"duration_ms":144401,"concrete_test":"Estimate the noise floor of the ε33 maps by selecting a 20×20 µm² defect-free region in Fig. 5(a), subtracting a low-order polynomial background, and computing the residual standard deviation σ. Then determine the peak lobe amplitude at each of the 39 dislocation sites and compare it with the local background distribution, for example requiring the lobe peak to exceed 3σ after masking the central 2 µm around each core. If the lobe amplitudes are not above 3σ, or if a similar lobe population is found in a defect-free control region, the ±(4 to 5)×10⁻⁵ strain-lobe result is not supported and the central strain-field claim should be downgraded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weakest load-bearing step is the quantitative claim that the axial-strain lobes have amplitudes of ±(4 to 5)×10⁻⁵ (abstract, Fig. 5, and conclusion). The paper itself reports that the axial strain in the surrounding matrix varies within approximately ±0.5×10⁻⁴, i.e. ±5×10⁻⁵, yet no noise floor, standard deviation, or confidence interval is given for the ε33 first-moment maps. A localized lobe of 4 to 5×10⁻⁵ is therefore not demonstrably above the background variability of the same map. The contrast could be a slowly varying artifact of the layer/substrate rocking-curve convolution (which the paper acknowledges for the tilt maps) or noise in the first-moment estimator. The dislocation-contrast simulations and Nye-tensor analysis that would independently support the lobe interpretation are relegated to supplementary material and cannot be checked from the preprint. This matters because the 'first maps of elastic strain around individual TDs' claim, and the strain-to-tilt decoupling that is the paper's main advance, rest on these lobes being real signal rather than background. By contrast, the reader's weakest assumption about the 7 µm KAM projection is less vulnerable: for both θ_B=5.605° (from 2θ=11.21°) and θ_B≈8.8°, t cosθ_B equals about 7 µm to within 1%, so the Bragg-angle discrepancy does not change the prong-separation conclusion.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports dark-field X-ray microscopy (DFXM) measurements in reflection geometry on a 7 µm HgCdTe/CdZnTe epilayer, claiming the first maps of individual threading-dislocation strain fields in this material system. The central claims are: (i) weak-beam difference images separate substrate-originated dislocation dots from elongated epilayer island features; (ii) kernel average misorientation (KAM) resolves each threading dislocation as a pair of prongs separated by ~7 µm, equal to the layer thickness and interpreted as the interface-emergence and surface-termination intersections of a single inclined line; (iii) axial strain lobes of approximately ±(4–5)×10⁻⁵ are localized within a few micrometres of dislocation cores; and (iv) strain and tilt are decoupled, with long-range tilt fields extending tens of micrometres. The analysis uses standard DFXM first-moment decomposition and compares against dislocation-theory estimates.","tokens_in":8904,"tokens_out":4918,"duration_ms":41999,"significance":"If the strain-lobe and KAM results hold, the paper would provide a valuable non-destructive, spatially resolved view of dislocation strain and rotation in an industrially relevant infrared-detector heterostructure. The manuscript has notable strengths: the KAM statistics are based on 39 dislocation pairs, the density estimate is compared with external etch-pit and near-field topograph values, and the interpretation is anchored in standard dislocation theory rather than a free-parameter fit. The weak-beam difference imaging idea is a genuinely useful experimental development. However, the central quantitative strain claim is not yet convincingly separated from the reported background variability, which is a load-bearing issue for the paper's main advance of strain–tilt decoupling.","major_comments":[{"comment":"The reported axial-strain lobes of ±(4–5)×10⁻⁵ are within the stated matrix background of approximately ±0.5×10⁻⁴ (i.e., ±5×10⁻⁵) in the same map. No noise floor, standard deviation, or confidence interval is given for the ε33 first-moment reconstruction, so the localized lobes are not demonstrably distinguishable from background variability. This is load-bearing because the claim of strain–tilt decoupling and the 'first maps of strain around individual TDs' rest on these lobes being real signal rather than noise or a rocking-curve convolution artifact (which the paper itself invokes to explain the wavy tilt background). The authors should quantify the noise in ε33 (e.g., from flat regions or repeated scans), show that the lobe amplitudes exceed it with statistical significance, or present a quantitative comparison with simulation that includes a background model.","section":"Fig. 5(a) and the paragraph beginning 'To examine the associated strain fields'"},{"comment":"The manuscript states that 'dislocation-contrast simulations for the three g-visible 1/2<110> variants reproduce the observed behaviour' and that 'the partition between edge-like and screw-like character is independently recovered from the Nye tensor components', but these analyses are only mentioned in a short paragraph and are not checkable from the preprint because the supplementary material is not provided. Since the alternating-sign multipolar strain pattern is atypical for a simple end-on edge dislocation, these simulations and Nye-tensor maps are essential verification. At minimum, a representative simulated strain map and a Nye-tensor map for one or two dislocations should be included in the main text or the supplementary material made available for review.","section":"Supplementary material (Nye-tensor analysis and dislocation-contrast simulations)"}],"minor_comments":[{"comment":"The text reports 2θ = 11.21° (implying θ_B ≈ 5.6°), but in the KAM geometry argument it uses θ_B ≈ 8.8°. Although both values give t cos θ_B ≈ 7 µm for t = 7 µm, this inconsistency should be reconciled by stating the correct Bragg angle and using it consistently throughout.","section":"Experimental section, 'Measurements'"},{"comment":"The sentence 'The dot-like and wavy line-like features that were unresolved in the millimetre-scale topographs of Ref. 20.' is an incomplete sentence; it should be revised to describe what is now resolved.","section":"Fig. 2 description"},{"comment":"The opening sentence of the conclusion is incomplete: 'we show that the elastic strain and lattice-rotation fields of individual TDs in an LPE-grown Hg1−xCdxTe epilayer.' It needs a verb such as 'are mapped' or 'are resolved'. The same paragraph also contains the awkward phrase 'an order of magnitude larger in angle', which should be clarified.","section":"Conclusion"},{"comment":"The text says the inner-edge opening median is 7.10 µm with σ_ln = 0.10, 'fixed at the 7µm epilayer thickness'. The wording is confusing: if the median is a fit parameter, it is not fixed at 7 µm; if it is fixed, the reported median of 7.10 µm is redundant. Please clarify whether the median was constrained to the thickness or fitted.","section":"Fig. 4 statistics paragraph"},{"comment":"Equation (1) uses Δd/d for the axial strain, while later text and figures use ε33. Define the equivalence explicitly to avoid ambiguity.","section":"General notation"}],"recommendation":"major_revision","confidential_remarks":"The supplementary material is essential to the central claims and should be made available to the reviewers; the current dependence on 'described but not shown' simulations is unsatisfying. The Bragg-angle inconsistency is a minor correctness issue but may indicate sloppy internal cross-checking. The strain-vs-background problem is the primary technical obstacle; if the authors can provide a noise floor and show statistical significance, the paper could be suitable for publication. The paper's scope fits a materials-science journal, though the letter format currently leaves too little room for methodological justification."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a genuine first — DFXM maps of individual threading dislocations in LPE HgCdTe/CdZnTe, separating substrate and epilayer defect families in a single weak-beam difference frame. That part I believe. The strain-field quantitative claim is softer than the abstract suggests.\n\nWhat's actually new: Ref. 19 used DFXM on an MCT detector but didn't resolve individual dislocation strain fields; here the KAM maps show each TD as a 7 µm prong pair, matching the epilayer thickness via t cosθ_B, and the dislocation density from 39 pairs agrees with prior etch-pit and near-field values. The weak-beam difference map separating dot-like substrate TDs from elongated layer features is a practical advance and looks convincing in the figures.\n\nThe soft spots are proportionate. Most important: the axial strain lobes of ±(4–5)×10⁻⁵ are claimed against a matrix background the paper itself reports as ±0.5×10⁻⁴. There is no noise floor, standard deviation, or confidence interval on the ε₃₃ first-moment maps, so the lobes are not yet distinguishable from background. I don't see a load-bearing flaw in the KAM interpretation: the Bragg angle text is inconsistent (2θ=11.21° gives θ_B≈5.6°, while the geometric argument uses ≈8.8°), but for both values t cosθ_B stays within 1% of 7 µm, so the prong-separation conclusion stands. The dislocation-contrast simulations and Nye-tensor analysis are relegated to supplementary material; given the strain-lobe issue, those should be in the main text or made available. Data availability only on request is also a limit, though common in this field.\n\nThe paper deserves a serious referee. The KAM and weak-beam results are solid, and the strain/tilt decoupling is the claimed advance — but the strain amplitude needs a proper noise characterization before it can be trusted as a quantitative result. I'd send it out with strong encouragement to tighten the statistics and fix the angle inconsistency.\n\nRecommendation: peer review, conditionally; the central measurement is likely right but the paper overstates its confidence in the strain lobes.","headline":"Real first maps of individual dislocation strain and rotation in MCT/CZT, but the quantitative strain lobes need a noise floor before they are convincing.","tokens_in":9571,"tokens_out":1669,"would_cite":true,"duration_ms":16542,"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":"Dark-field X-ray microscopy maps the elastic strain and lattice rotation of individual threading dislocations in a HgCdTe/CdZnTe epilayer.","keywords":["dark-field X-ray microscopy","threading dislocations","HgCdTe/CdZnTe","elastic strain mapping","lattice rotation","kernel average misorientation","infrared detector materials"],"falsifier":"Re-measure the same dislocation field with a different Bragg reflection or with the sample rotated azimuthally, and check whether the KAM prong separation scales as $t\\cos\\theta_B$ using the actually measured diffraction angle; if the separation stays constant or scales differently, the thickness interpretation collapses. A direct cross-check is to compare DFXM prong positions with atomic-force-microscopy step distortions at the surface and with a cross-sectional transmission electron micrograph of the same dislocation line.","tokens_in":8404,"feed_emoji":"🔬","tokens_out":10658,"duration_ms":90732,"temperature":0.7,"pith_summary":"The paper seeks to establish that dark-field X-ray microscopy can separate, for individual threading dislocations in a HgCdTe/CdZnTe epilayer, the elastic strain they carry from the lattice rotation they cause, a separation that single rocking-curve topography cannot make. Working with a 7-micrometre-thick (111) epilayer in reflection geometry, the authors produce maps in which each dislocation shows compact axial-strain lobes of roughly $\\pm(4\\text{ to }5)\\times10^{-5}$ within a few micrometres of the core and a lattice-rotation field extending over tens of micrometres. They also show that weak-beam images on the two flanks of the rocking curve, subtracted from one another, separate dot-like substrate threading dislocations from the elongated in-plane island defects they nucleate in the layer. This matters because these dislocations set the minority-carrier lifetime and dark-current floor of mid-wave infrared focal-plane arrays, so knowing where strain actually sits is a step toward controlling it.","feed_headline":"X-ray maps strain and rotation at single dislocations","feed_subtitle":"Dark-field X-ray microscopy separates strain from lattice tilt, pinning the defect footprint that sets detector dark current.","key_machinery":"The carrying mechanism is dark-field X-ray microscopy (DFXM): an X-ray objective lens placed in the diffracted beam forms a magnified image of the illuminated volume and acts as a pinhole in reciprocal space. In this geometry, a rocking scan in the sample angle $\\phi$ at fixed $2\\theta$ maps lattice rotation, while a scan in $2\\theta$ at fixed $\\phi$ maps axial strain, and the first moments of the two scans give the $\\epsilon_{33}$ and tilt maps. The paper's additional step is to record weak-beam images on the two flanks of the convolved layer-plus-substrate rocking curve and take their difference, so that substrate-originated dots and epilayer island features appear with opposite or same sign in a single frame. Kernel average misorientation (KAM), a local magnitude of the lattice-orientation gradient, then turns each dislocation into a prong pair whose separation is read through the geometric projection $t\\cos\\theta_B\\approx t$ for a near-normal line.","core_discovery":"The central claim is that a threading dislocation crossing a 7-micrometre HgCdTe epilayer leaves a compact, alternating-sign axial-strain signature within about 5 micrometres of its core and a long-range lattice-rotation field roughly an order of magnitude larger in angular spread, and that both can be imaged in one reflection-geometry dark-field X-ray microscopy measurement. By scanning the sample angle at fixed diffraction angle and the diffraction angle at fixed sample angle, the authors separate lattice rotation from axial strain and reconstruct $\\epsilon_{33}$ and tilt maps. The kernel average misorientation maps resolve each dislocation as a pair of prongs separated by about 7 micrometres, which they identify with the dislocation crossing the full film, one prong at the buried interface and one at the free surface. The measured strain amplitudes agree with a $\\tfrac{1}{2}\\langle110\\rangle$ glide dislocation estimate, and the lobe polarity is linked to the Burgers vector and the mixed edge-screw character of the line. The authors present this as the first mapping of individual dislocation strain and rotation fields in the MCT/CZT system.","pith_inferences":["The image-depth interpretation can be tested without destroying the sample: re-imaging the same region with a different Bragg reflection, or after azimuthal rotation, should shift the KAM prong separation according to $t\\cos\\theta_B$; if the separation is invariant or scales differently, the surface/interface assignment would need revision.","The alternating-sign multipolar strain pattern around each core suggests a route to infer the local Burgers vector orientation and its depth-dependent edge-screw mix from DFXM maps alone, which the authors do not develop quantitatively.","The weak-beam difference scheme should transfer to other lattice-mismatched II-VI heterostructures and to device-grade arrays, giving a fast non-destructive screen for strain hot spots before detector processing.","Connecting these single-dislocation strain maps to dark-current or lifetime maps on the same wafer would test the implicit claim that the strain field, not the core, sets the electrically active defect footprint."],"forward_implications":["Each threading dislocation's electrically active footprint is set by its strain field, which reaches micrometres to tens of micrometres, rather than by the nanometre-scale core alone, so local band-gap fluctuations from the measured strain may act as recombination and trapping sites.","The 7-micrometre prong separation and its narrow lognormal spread show that KAM maps give a direct, geometry-controlled readout of epilayer thickness at each dislocation.","Weak-beam difference imaging separates substrate threading dislocations from epilayer nucleation features in one frame, removing the need to compare separately reconstructed maps to see correlated defects.","Tilt domains spanning 50 to 100 micrometres, comparable to a focal-plane pixel, imply that a single dislocation network can produce pixel-to-pixel non-uniformity in responsivity and dark current.","Because strain and tilt are decoupled, the edge-like versus screw-like character of each dislocation can be partitioned and connected to spiral-growth handedness and to the bending of dislocations into in-plane misfit segments."],"supporting_citations":[{"why":"Supplies the same LPE sample's prior near-field topography, the 7-micrometre thickness, the lattice mismatch, and the dot/island feature catalogue that the DFXM maps reproduce.","marker":"20"},{"why":"Gives the standard DFXM decomposition used to reconstruct axial strain and tilt from (phi, 2theta) scans.","marker":"12"},{"why":"Previous DFXM strain and rotation maps of individual misfit dislocations in epitaxial BiFeO3; the through-thickness superposition effect is used to interpret the alternating strain lobes.","marker":"15"},{"why":"Introduces dark-field X-ray microscopy and its reciprocal-space filtering by an objective lens in the diffracted beam.","marker":"11–13"},{"why":"Documents liquid-phase-epitaxy growth of MCT/CZT and the Zn-fraction lattice-match tuning that sets the epitaxial strain context.","marker":"5"},{"why":"Provides the elastic displacement estimate and dislocation strain fields used to compare the measured strain amplitudes with half-<110> glide dislocations.","marker":"26,27"},{"why":"Shows electron-channelling contrast lobe geometry in GaN, used as an analogue for interpreting axial-strain lobe polarity around MCT threading dislocations.","marker":"28"}],"fun_headline_variants":["X-ray separates lattice tilt from strain at one dislocation","Single-dislocation strain and rotation fields resolved in HgCdTe","Dark-field X-ray pinpoints strain at threading dislocations","Measuring individual dislocation strain in detector-grade HgCdTe","One dislocation, two fields: strain and rotation mapped"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the geometric identification of the two KAM prongs with a single dislocation crossing the whole film: it assumes a near-normal dislocation line of length equal to the 7-micrometre layer thickness projects to a prong separation of $t\\cos\\theta_B$, and the paper applies this with $\\theta_B\\approx8.8^\\circ$ even though the stated diffraction geometry gives $\\theta_B\\approx5.6^\\circ$, so if that projection is wrong, the reading of each prong pair as surface and interface intersections falls.","fun_headline_variants_meta":{"raw":{"variants":["X-ray separates lattice tilt from strain at one dislocation","Single-dislocation strain and rotation fields resolved in HgCdTe","Dark-field X-ray pinpoints strain at threading dislocations","Measuring individual dislocation strain in detector-grade HgCdTe","One dislocation, two fields: strain and rotation mapped"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000707,"raw_usage":{"total_tokens":3216,"prompt_tokens":1007,"completion_tokens":2209,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":623,"completion_tokens_details":{"reasoning_tokens":2129}},"tokens_in":623,"tokens_out":2209,"duration_ms":15458,"temperature":1.0,"reasoning_tokens":2129,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T05:37:30.721851+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-measure the same dislocation field with a different Bragg reflection or with the sample rotated azimuthally, and check whether the KAM prong separation scales as $t\\cos\\theta_B$ using the actually measured diffraction angle; if the separation stays constant or scales differently, the thickness interpretation collapses. A direct cross-check is to compare DFXM prong positions with atomic-force-microscopy step distortions at the surface and with a cross-sectional transmission electron micrograph of the same dislocation line.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the same LPE sample's prior near-field topography, the 7-micrometre thickness, the lattice mismatch, and the dot/island feature catalogue that the DFXM maps reproduce."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the standard DFXM decomposition used to reconstruct axial strain and tilt from (phi, 2theta) scans."},{"cited_title":"Simons and A","cited_arxiv_id":null,"evidence_quote":"Previous DFXM strain and rotation maps of individual misfit dislocations in epitaxial BiFeO3; the through-thickness superposition effect is used to interpret the alternating strain lobes."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents liquid-phase-epitaxy growth of MCT/CZT and the Zn-fraction lattice-match tuning that sets the epitaxial strain context."},{"cited_title":"Pelliciari , title =","cited_arxiv_id":null,"evidence_quote":"Shows electron-channelling contrast lobe geometry in GaN, used as an analogue for interpreting axial-strain lobe polarity around MCT threading dislocations."}],"review_version":1}