{"id":"a13d5fc2-3e3a-46c9-b519-e962af150a28","arxiv_id":"2509.06242","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Liquid phase epitaxy yields strained YIG films on diamagnetic YSGG with FMR linewidths below 1 mT at 3 K, avoiding the paramagnetic substrate losses of YIG/GGG.","lead":"Researchers grew ultrathin yttrium iron garnet films on a non-magnetic garnet substrate and measured their spin-wave losses down to 3 kelvin. The films retain low losses at cryogenic temperatures, opening a route to quantum devices that use spin waves.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No composition depth profile for YIG/YSGG; cation interdiffusion could introduce extrinsic low-temperature linewidth contributions.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: the absence of a composition depth profile for the YIG/YSGG interface. This is the most critical gap because the central claim relies on the film being pure YIG to attribute the low linewidth solely to the diamagnetic substrate. The paper provides strong structural evidence (RSM, Laue oscillations) but no chemical evidence. The proposed ToF-SIMS test directly addresses this gap. Since the reader already conditioned acceptance on this issue, my read does not change the verdict; it reinforces the condition.","tokens_in":32227,"tokens_out":4462,"duration_ms":49033,"concrete_test":"Perform time-of-flight secondary ion mass spectrometry (ToF-SIMS) depth profiling on YIG/YSGG sample A, monitoring Sc+ and Ga+ counts as a function of depth from the film surface through the interface. If the concentration of Sc and Ga in the YIG layer is below 0.1 at% (or the detection limit) throughout the film thickness, the interdiffusion concern is resolved; if significant diffusion is present, the measured FMR linewidth must be re-evaluated as including extrinsic contributions.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that LPE-grown YIG on commercial YSGG substrates achieves FMR linewidth below 1 mT at 3 K, attributed to the diamagnetic substrate. This assumes the film remains stoichiometric YIG with negligible interdiffusion of Sc or Ga from the substrate into the film. XRD/RSM only confirm structural coherence (fully strained, Laue oscillations) and are insensitive to dilute cation substitution in the first few nanometers of the interface. The authors cite Refs. [23,24] for minimal interdiffusion in LPE YIG/GGG, but YSGG has a different lattice mismatch and chemical composition; no SIMS/EDX/TEM composition depth profile is provided. If Sc or Ga substitutes into YIG, the magnetization, anisotropy, and low-temperature relaxation would be modified, meaning the measured linewidth would not reflect the intrinsic properties of YIG, and the claim that YSGG is a suitable diamagnetic substrate for low-loss YIG would be undermined. This is the weakest link in supporting the central claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports liquid-phase-epitaxial growth of sub-100-nm YIG films on (111)-oriented yttrium scandium gallium garnet (YSGG) substrates, with X-ray reflectivity, diffraction, and reciprocal-space-map evidence for fully strained, pseudomorphic growth below roughly 90–110 nm. Broadband FMR measurements from 293 K down to 3 K compare YIG/YSGG with YIG/GGG. On YSGG, the FMR linewidth remains nearly flat down to 3 K; the best OOP value is 0.7 mT at 46 GHz, with a conservative upper bound α<2×10−4. The large low-temperature linewidth increase of YIG/GGG is attributed primarily to inhomogeneous stray fields from the paramagnetic GGG substrate, modeled in Appendix C. The authors conclude that LPE YIG on commercial YSGG removes the substrate-paramagnetism bottleneck for cryogenic magnonics.","tokens_in":32524,"tokens_out":8469,"duration_ms":89370,"significance":"If substantiated, the result is significant: it provides a straightforward LPE route to low-loss iron-garnet films on a diamagnetic substrate for cryogenic magnonics and hybrid quantum devices. The strengths are real: linewidths are directly measured with documented broadband FMR and field-differential analysis; apparent Gilbert damping is separated from inhomogeneous broadening; the GGG stray-field contribution is modeled quantitatively with an explicit procedure; and the claims are qualified as applying to unpatterned films and as upper bounds. The principal gap is chemical characterization of the film and of the film/substrate interface, which is important for attributing the measured low linewidth to intrinsic YIG rather than to a fortuitous substituted-garnet composition.","major_comments":[{"comment":"The central attribution of the low-temperature linewidth to intrinsic YIG on a diamagnetic substrate rests on the assumption that the LPE film is stoichiometric YIG with negligible Sc/Ga interdiffusion at the YIG/YSGG interface. The X-ray RSM and Laue-oscillation data establish structural coherence and full strain but cannot detect dilute cation substitution. Refs. [23,24] document minimal interdiffusion only for LPE YIG/GGG; YSGG has a different lattice mismatch, composition, and growth chemistry. If Sc or Ga enters the film, the magnetization, anisotropy, and low-temperature relaxation would be modified, so the measured linewidth would not exclusively reflect intrinsic YIG properties. Please provide a composition depth profile (SIMS, EDX, or TEM) or explicitly qualify the conclusions as applying to the actual film composition rather than to intrinsic YIG.","section":"Section II, Table I and Fig. 2"},{"comment":"The quantitative claims—linewidth below 1 mT at 3 K, α<2×10−4, and the weak temperature dependence—are reported without uncertainties or error bars. The fits to μ0ΔH(f) are shown, but the fit parameters in Table I and Figs. 5, 7, and 9 have no confidence intervals. Since the comparison between YSGG A and B and the threshold 'below 1 mT' are central to the paper, representative uncertainty estimates (at least for the headline values and one full temperature sweep) are needed to support those claims.","section":"Table I and Figs. 5, 7, 9"}],"minor_comments":[{"comment":"Several figure captions and Appendix C equations contain garbled placeholder text (e.g., '/uni...' sequences). The final version should render all symbols cleanly.","section":"Fig. 1 caption; Appendix C"},{"comment":"The label 'no subs.' in Figs. 5(c) and 5(d) is not defined in the caption. Please define it as 'after subtracting the substrate stray-field contribution ΔB_stray'.","section":"Fig. 5"},{"comment":"The notation for the stray-field broadening is inconsistent: 'ΔB_stray', 'ΔBstray', and 'ΔB stray' are used. Please unify.","section":"Throughout"},{"comment":"The text states that fits exclude low-frequency data points below 6 GHz, but the exact cutoff used for each sample and temperature is not given. Please state the cutoff, since Table I parameters depend on it.","section":"Section II, FMR analysis"}],"recommendation":"major_revision","confidential_remarks":"The experimental dataset is valuable and the central linewidth measurement is credible. I would not reject the paper; the main obstacles are the missing composition depth profile and the absence of uncertainty quantification. If the authors can supply a SIMS/EDX/TEM profile or reframe the claims to avoid the intrinsic-YIG attribution, and add representative error bars, the paper would be acceptable. The self-citation in Appendix C is innocuous and does not affect the YSGG central result."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main thing you should know: this is the first LPE growth of YIG on YSGG, and the headline result—FMR linewidth below 1 mT at 3 K in unpatterned films—is directly measured, not inferred from fits. That is a genuinely useful step for cryogenic magnonics, and it goes beyond the earlier sputtered YIG/YSGG work. The strain-relaxation threshold around 100 nm and the full 3–300 K FMR comparison with GGG are new data, and the paper treats them carefully.\n\nWhat the paper does well: the FMR analysis is clean. The authors separate apparent Gilbert damping from inhomogeneous broadening, model the GGG stray-field contribution using a Brillouin function and demagnetizing factors, and show that most of the low-temperature linewidth rise in YIG/GGG is extrinsic. They also qualify their own result honestly—unpatterned films, an upper bound on alpha, and a clear statement that strain and roughness matter. The X-ray data (RSM, Laue oscillations) support the claim of fully strained, high-quality growth.\n\nSoft spots: the stress-test concern about interdiffusion is fair but not damning. There is no SIMS/TEM composition profile across the YIG/YSGG interface, so dilute Sc or Ga substitution cannot be strictly excluded. But the central claim is a measured linewidth, not a claim of compositional purity. Even if some interdiffusion exists, the linewidth is still below 1 mT at 3 K, and the diamagnetic substrate advantage over GGG is demonstrated by direct comparison. I would raise the missing profile as a request in review, not as a blocker. The lack of error bars on the linewidth values is a minor issue—standard practice in FMR papers, but worth asking for.\n\nThe citation pattern is fine. The only self-citation is to their own field-differential FMR method, which is a methodological reference, not a circular load-bearing claim.\n\nBottom line: this is a useful, reproducible experimental contribution. It is not a conceptual breakthrough—the idea of replacing GGG with a diamagnetic substrate predates it—but it adds the highest-quality growth method to that effort and reports clean cryogenic data. I would send this to peer review without hesitation, and I would cite it in my own work.\n\nRecommendation: accept for review, with the composition depth profile and error bars as requested revisions.","headline":"First LPE growth of strained YIG on diamagnetic YSGG delivers cryogenic FMR linewidths below 1 mT at 3 K; a solid, honest experimental paper that deserves peer review, with a missing composition depth profile as the main (non-fatal) gap.","tokens_in":32971,"tokens_out":1545,"would_cite":true,"duration_ms":19434,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["76.50.+g","75.70.-i"],"model":"deepseek-v4-flash","headline":"Replacing the standard paramagnetic GGG substrate with commercially available diamagnetic YSGG lets liquid-phase-epitaxy yttrium iron garnet films keep ferromagnetic resonance linewidths below 1 mT down to 3 K, removing the main obstacle to","keywords":["yttrium iron garnet","liquid phase epitaxy","ferromagnetic resonance","cryogenic magnonics","diamagnetic substrate","yttrium scandium gallium garnet","Gilbert damping","strain relaxation"],"falsifier":"A secondary-ion mass spectrometry or EDX depth profile across the YIG/YSGG interface: measurable scandium or gallium in the YIG layer beyond the first nanometre, or at levels above roughly 0.01 per formula unit, would mean the 3 K linewidth carries extrinsic impurity losses beyond those the paper assigns to strain. Alternatively, growing the same film on a strain-compensated or lattice-matched diamagnetic substrate and seeing the 3 K linewidth drop below 0.7 mT would confirm that strain, not something intrinsic to YSGG, sets the current floor.","tokens_in":32210,"feed_emoji":"🧲","tokens_out":6059,"duration_ms":62970,"temperature":0.7,"pith_summary":"This paper tries to establish that thin yttrium iron garnet (YIG) films can stay as low-loss at 3 K as they are at room temperature, provided they are grown on the right substrate. The key move is replacing the standard gadolinium gallium garnet (GGG) substrate — strongly paramagnetic and a known killer of magnon lifetimes at low temperatures — with commercially available, diamagnetic yttrium scandium gallium garnet (YSGG). Using liquid phase epitaxy, the authors grow fully strained YIG films under 100 nm thick on YSGG and measure ferromagnetic resonance linewidths below 1 mT at 3 K, with almost no temperature dependence between 3 and 300 K. A comparison film on GGG degrades sharply at low temperatures, and the paper shows quantitatively that most of that degradation is a stray-field artifact of the GGG substrate itself. If correct, this gives cryogenic magnonics and quantum-hybrid experiments a straightforward thin-film platform without lift-off or substrate-removal steps.","feed_headline":"Diamagnetic substrate keeps YIG film linewidth under 1 mT at 3 K","feed_subtitle":"Swapping GGG for YSGG removes the paramagnetic-substrate losses that block cryogenic magnonics.","key_machinery":"The central object is the YSGG substrate (yttrium scandium gallium garnet), a commercial diamagnetic garnet whose lattice parameter sits about 0.7 percent below YIG's, close enough to support fully strained pseudomorphic YIG films up to roughly 100 nm thick. Its job is to eliminate the paramagnetic Gd3+ moments that, in the standard GGG case, couple to the film and broaden the resonance at cryogenic temperatures. The argument's quantitative engine is a two-part model: a mean-field Brillouin-function description of GGG's magnetization, and a computed stray-field distribution (using demagnetizing factors and field-profile statistics) that yields the extra linewidth contribution subtracted from","core_discovery":"On its own terms: growing YIG by liquid phase epitaxy on a diamagnetic YSGG substrate removes the dominant low-temperature loss channel in thin-film magnonics. The paper reports a ferromagnetic resonance linewidth of 0.7 mT at 3 K and 46 GHz (below 1 mT across the range), a Gilbert damping upper bound of 2 × 10^-4, and a linewidth that barely moves between 3 and 100 K, in both in-plane and out-of-plane field geometries. For comparison, YIG/GGG films measured in the same setup show a strong low-temperature linewidth increase, which the authors trace, via a shape-dependent stray-field model, to the paramagnetic GGG substrate's magnetization rather than to the YIG film itself. The residual temp","pith_inferences":["If strain is the dominant residual loss source, lattice-matched diamagnetic substrates or strain-compensating buffer layers should push 3 K linewidths toward the bulk-YIG limit, possibly extending usefulness toward millikelvin temperatures.","The stray-field correction model implies that some previously reported low-temperature damping values for YIG/GGG were substrate artifacts; re-analyzing such data with comparable corrections would revise the apparent temperature dependence of the Gilbert damping.","Patterning these films into microstructures should remove most inhomogeneous broadening, potentially revealing an intrinsic damping below 1 × 10^-4 at 3 K and enabling a direct comparison with bulk YIG spheres.","The same LPE growth on YSGG should transfer to substituted iron garnets, allowing strain-tuned anisotropy and magnetization while keeping cryogenic compatibility."],"forward_implications":["Cryogenic magnonic devices can use unpatterned LPE YIG films on commercial YSGG directly, with no need to lift the film off its growth substrate.","The FMR linewidth stays nearly flat from 300 K down to 3 K, so the paramagnetic-substrate degradation that dominates YIG/GGG below about 50 K is avoided in both in-plane and out-of-plane geometries.","The fully strained thickness window (below roughly 100 nm) defines the usable range for YIG-on-YSGG films; above it, strain relaxation sharply degrades the linewidth.","The measured upper bound on intrinsic Gilbert damping (below 2 × 10^-4 at 3 K) is low enough to support coherent magnonic and hybrid-superconducting-qubit experiments at cryogenic temperatures.","Residual inhomogeneous broadening is tied to strain and substrate quality, pointing to strain mitigation as the next improvement step for these films."],"supporting_citations":[{"why":"Establishes the temperature-dependent magnetic properties of ultrathin LPE YIG and underpins the assumption of minimal cation interdiffusion in LPE-grown films.","marker":"[23]"},{"why":"Shows negligible magnetic losses at low temperatures in LPE-grown YIG films, the basis for expecting the interface to stay chemically inert on the new substrate.","marker":"[24]"},{"why":"Prior demonstration of low-damping YIG on YSGG at 2 K via sputtering, the baseline this work improves on with LPE growth.","marker":"[30]"},{"why":"Provides lattice-tunable substituted iron garnets as the context for mitigating strain on diamagnetic substrates.","marker":"[31]"},{"why":"Supplies the GGG substrate stray-field model used to compute and subtract the stray-field contribution to the YIG/GGG linewidth.","marker":"[21]"},{"why":"Quantifies the damping enhancement in YIG at millikelvin temperatures due to the GGG substrate, the basis for the delta-B_stray linewidth contribution formula.","marker":"[37]"},{"why":"Provides the bulk-YIG low-temperature linewidth reference and the few-parts-per-million rare-earth impurity estimate used to interpret the 30–40 K linewidth peak.","marker":"[14]"},{"why":"Documents the increased low-temperature damping in YIG thin films on GGG, the specific problem the paper claims to solve.","marker":"[18]"},{"why":"Establishes the low-damping quality achievable in thin LPE YIG films, the growth-method foundation of the present work.","marker":"[35]"}],"fun_headline_variants":["YIG on YSGG: cryo-ready magnonics with sub-1mT linewidth","LPE-grown YIG on diamagnetic substrate tames low-T losses","Swap GGG for YSGG: YIG films stay loss-free at 3K","Cryo-friendly YIG: LPE on YSGG keeps linewidth under 1 mT"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The central claim assumes that the YIG/YSGG interface is as chemically inert as the YIG/GGG interface — that scandium and gallium from the substrate do not diffuse into the film during growth — yet the paper supports this only with X-ray structural data and no direct composition depth profile.","fun_headline_variants_meta":{"raw":{"variants":["YIG on YSGG: cryo-ready magnonics with sub-1mT linewidth","LPE-grown YIG on diamagnetic substrate tames low-T losses","Swap GGG for YSGG: YIG films stay loss-free at 3K","Cryo-friendly YIG: LPE on YSGG keeps linewidth under 1 mT"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000487,"raw_usage":{"total_tokens":2269,"prompt_tokens":805,"completion_tokens":1464,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":549,"completion_tokens_details":{"reasoning_tokens":1368}},"tokens_in":549,"tokens_out":1464,"duration_ms":10692,"temperature":1.0,"reasoning_tokens":1368,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T23:50:58.147161+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A secondary-ion mass spectrometry or EDX depth profile across the YIG/YSGG interface: measurable scandium or gallium in the YIG layer beyond the first nanometre, or at levels above roughly 0.01 per formula unit, would mean the 3 K linewidth carries extrinsic impurity losses beyond those the paper assigns to strain. Alternatively, growing the same film on a strain-compensated or lattice-matched diamagnetic substrate and seeing the 3 K linewidth drop below 0.7 mT would confirm that strain, not something intrinsic to YSGG, sets the current floor.","supporting_citations":[{"cited_title":"Beaulieu, N","cited_arxiv_id":null,"evidence_quote":"Establishes the temperature-dependent magnetic properties of ultrathin LPE YIG and underpins the assumption of minimal cation interdiffusion in LPE-grown films."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows negligible magnetic losses at low temperatures in LPE-grown YIG films, the basis for expecting the interface to stay chemically inert on the new substrate."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior demonstration of low-damping YIG on YSGG at 2 K via sputtering, the baseline this work improves on with LPE growth."},{"cited_title":"Legrand, Y","cited_arxiv_id":null,"evidence_quote":"Provides lattice-tunable substituted iron garnets as the context for mitigating strain on diamagnetic substrates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the GGG substrate stray-field model used to compute and subtract the stray-field contribution to the YIG/GGG linewidth."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Quantifies the damping enhancement in YIG at millikelvin temperatures due to the GGG substrate, the basis for the delta-B_stray linewidth contribution formula."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the bulk-YIG low-temperature linewidth reference and the few-parts-per-million rare-earth impurity estimate used to interpret the 30–40 K linewidth peak."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Documents the increased low-temperature damping in YIG thin films on GGG, the specific problem the paper claims to solve."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the low-damping quality achievable in thin LPE YIG films, the growth-method foundation of the present work."}],"review_version":1}