{"id":"e8df3f4a-8242-4fbc-a5bb-a46a4135862e","arxiv_id":"1908.06536","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"MnGa films are chemically stable in acetone and ethanol, but are violently attacked by deionized water and acid solutions, and TMAOH developer with prolonged exposure.","lead":"MnGa magnetic films survive acetone and ethanol but are quickly damaged by water and acid solutions, while oxygen plasma only oxidizes the surface. This study gives device makers a practical list of which fabrication steps to avoid when working with MnGa spintronic devices.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central claim that DI water and acids attack MnGa violently rests on uncalibrated resistance changes and qualitative microscopy; a thickness or composition measurement is needed to validate resistance as a proxy for attack.","rationale":"The reader's weakest assumption correctly identifies the resistance proxy as the load-bearing point. My independent reading reaches the same conclusion: Figure 3 is the only quantitative evidence for the DI water and TMAOH claims, and it lacks validation, replicates, and uncertainty estimates. The acid claim is supported by microscopy but not quantified. The paper does provide a useful qualitative ranking, and the stability in acetone and ethanol is plausible and consistent with resistance data. An additional calibration experiment is needed to confirm that resistance reflects material loss rather than contact or surface effects; without it, the stated processing windows ('quick rinse', 'development <1 min is safe') are not quantitatively justified. This does not refute the paper's overall direction, so the conditional verdict remains appropriate.","tokens_in":4613,"tokens_out":2903,"duration_ms":37204,"concrete_test":"Fabricate fresh 30 nm MnGa/MgO samples and soak separate pieces in DI water, TMAOH developer, acetone, and ethanol for 0, 1, 3, 5, and 10 min. After the soak, evaporate Ti/Au contacts (so contacts are never exposed to the chemicals) and measure sheet resistance using a four-probe van der Pauw geometry. Independently measure film thickness by X-ray reflectivity and Mn/Ga composition by XPS depth profiling on the same samples. If resistance increases track measurable thickness loss or Mn/Ga depletion within the same time window, the proxy is validated; if resistance changes without material loss, the conclusions should be revised to surface oxidation or contact effects.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 2.3 and Figure 3, the paper uses the electrical resistance of 30 nm MnGa films with a 1.5 nm MgO cap as the sole quantitative measure of chemical attack. A resistance increase could equally arise from contact degradation, surface oxidation, or delamination rather than from material loss, yet the conclusions are phrased in terms of violent etching and safe processing windows (e.g., TMAOH development under 1 min is stated to be safe). No calibration against film thickness, composition, or surface morphology is provided, and no error bars or replicate measurements are shown. The acid results in Section 2.2 are only qualitatively described from optical microscopy, so the quantitative severity of acid attack is not established. The chemical conditions (acid concentrations, temperature, water resistivity) are also unspecified, making the reported rates difficult to reproduce. Because the practical recommendations (avoid water and acids, limit TMAOH exposure) depend on this unvalidated proxy, the central claim is not fully supported as stated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental study of the chemical stability of 30 nm L10-ordered MnGa films, capped with 1.5 nm MgO and grown on GaAs, in contact with acids (HCl, H3PO4, H2SO4), acetone, ethanol, deionized water, TMAOH developer, and oxygen plasma. The authors conclude that acetone and ethanol are safe; deionized water and acids attack the films severely; TMAOH attack is slow, with development times below 1 min considered safe; and oxygen plasma oxidizes and passivates the surface. The conclusions are based on optical microscopy of wet-etched patterns (Section 2.2) and on resistance-versus-time measurements of unpatterned films (Section 2.3). The authors recommend avoiding water and acids in MnGa device fabrication whenever possible and provide guidance on TMAOH development and oxygen plasma exposure.","tokens_in":4754,"tokens_out":3423,"duration_ms":34111,"significance":"If the conclusions are supported, the paper would provide practically important processing guidelines for MnGa-based spintronic devices, a topic for which published chemical-stability data are scarce. The systematic comparison of standard cleanroom chemicals, the use of a well-characterized Mn50Ga50 film with XPS verification, and the explicit caution about water exposure are useful contributions. The practical recommendations (avoid water and acids, limit TMAOH development time, use oxygen plasma for passivation) are falsifiable and directly relevant to device integration. However, the significance is conditional: the central quantitative evidence rests on an uncalibrated resistance proxy and on qualitative microscopy, so the strength of the practical claims is not yet established. The paper does not include replicates, error bars, or validated measures of material loss, which limits its current value as a quantitative reference.","major_comments":[{"comment":"The central quantitative evidence is the change in resistance of the 30 nm MnGa film during soaking. The paper does not establish that resistance is a faithful proxy for chemical attack. An increase in resistance could result from loss of conductive cross-section, but equally from surface oxidation, delamination, or degradation of the electrical contact between the probes and the film. No calibration is provided against film thickness loss (e.g., XRR, TEM, or profilometry), composition change (e.g., XPS before and after), or surface morphology (e.g., AFM). Without such validation, the quantitative statements such as \"R reaches 20 MΩ in 5 minutes\" in deionized water and the claim that TMAOH development under 1 min is safe are not fully supported. Please add a direct measure of attack for at least one representative condition (e.g., DI water and TMAOH) and describe the measurement geometry, contact method, and number of replicates.","section":"Section 2.3, Figure 3"},{"comment":"The acid etching study is presented only qualitatively. The text states that HCl, H3PO4, and H2SO4 solutions produce strong sidewall etching \"regardless of the combination of different volume ratio and etching time,\" but no concentrations, temperatures, or etching times are listed, and no quantitative etch rate, undercut distance, or reproducibility data are shown. The optical images in Figure 2(c) are evocative but cannot support the severity ranking between acids or the statement that wet etching is non-uniform without controlled experiments. Please specify the acid concentration and temperature for each test and provide quantitative measures of attack (e.g., step height or sheet resistance change) in addition to optical images.","section":"Section 2.2, Figure 2"},{"comment":"The chemical conditions are under-specified, which limits reproducibility. The developer is described as \"TMAOH solution (TMAOH:H2O=4:1)\" in Section 2.2 but as \"75% TMAH developer\" in Section 2.3; the deionized water resistivity is not given; the oxygen plasma process is specified only as \"power of 200 W\" with no pressure, gas flow, or time dependence of the passivation effect. In addition, the resistance measurement geometry in Figure 3(a) is not described in the text (two-point vs four-point, probe arrangement, contact stability). Please report these parameters so that the soaking and plasma conditions can be reproduced.","section":"Section 2.3 and Section 2.2"},{"comment":"All studied films have a 1.5 nm MgO capping layer, so the measured resistance changes reflect attack of the MgO/MnGa stack rather than of MnGa alone. The interpretation of the resistance increase in water and TMAOH could be influenced by pinholes, edge attack, or delamination of the cap, and the conclusions are phrased as intrinsic MnGa stability. Please discuss the role of the MgO cap or provide measurements on films without a cap or with different cap thicknesses to separate cap effects from the intrinsic chemical stability of MnGa.","section":"Section 2.3, sample structure"}],"minor_comments":[{"comment":"The phrase \"has unsettled\" is grammatically incorrect; consider \"has remained unsettled\" or \"is not settled.\"","section":"Abstract"},{"comment":"The phrase \"we first exam\" should read \"we first examine.\"","section":"Section 2.2, paragraph 1"},{"comment":"Figure 3(b) shows curves without error bars or replicate symbols; please state whether each curve is from a single sample and whether multiple samples gave consistent results.","section":"Section 2.3, Figure 3"},{"comment":"The XPS composition verification is only described as \"asymmetric elemental XPS peaks\" and a stated Mn:Ga ratio of 1:1; please report the quantified atomic percentages obtained with the stated sensitivity factors so the reader can verify the composition measurement.","section":"Section 2.1, Figure 1(b)"},{"comment":"The statement that oxygen plasma \"can passivate the MnGa surface\" is based on a resistance change after 1 min, but the resistivity or sheet resistance is not reported; please state the magnitude and sign of the resistance change and clarify whether the passivation was confirmed by any surface analysis (e.g., XPS of the oxide layer).","section":"Section 2.3, oxygen plasma"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is concise and addresses a practical topic, but the evidence base is currently too thin for a full research paper. The requested calibration of the resistance proxy and quantitative acid-etch data are essential before publication. I would be satisfied with a revised version that adds thickness or composition validation for the resistance measurements, specifies all chemical and plasma conditions, and qualifies the conclusions accordingly."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What's actually new here is the first systematic test of L10-MnGa against the chemicals a device fabricator would actually use: acids, acetone, ethanol, DI water, TMAOH developer, and oxygen plasma. That fills a real gap. The paper is straightforward, readable, and does not overreach in its main qualitative conclusions. The XPS composition check is good practice, and the acid-etching result is visually unambiguous: resist-covered wires vanish while uncovered regions remain, which speaks for itself as evidence of severe sidewall attack.\n\nThe soft spots are real but not fatal. The quantitative claims rest entirely on resistance changes of a 30 nm MnGa film with a 1.5 nm MgO cap, and that proxy is never validated. A resistance jump to 20 MΩ in DI water could be material loss, but it could equally be contact degradation, oxidation, or delamination. There are no thickness or composition measurements after soaking, no error bars, no replicates. The acid data are only qualitative optical images, and the acid concentrations and temperatures are not specified. The TMAOH \"under 1 min is safe\" conclusion follows from resistance being stable for 3 minutes, but without knowing what resistance change is tolerable for a real device, that quantitative window is shaky.\n\nThat said, the central practical message is almost certainly right: water and acids should be avoided, acetone and ethanol are safe, and TMAOH exposure should be kept short. Even if the mechanism behind the resistance change is not pinned down, the paper is honest about what it measured and does not pretend to explain chemistry that was not studied.\n\nWho gets value from this: anyone fabricating MnGa-based spintronic devices—magnetic tunnel junctions, sensors, oscillators—who currently has to guess whether a standard cleaning or lithography step will destroy the film. For that audience, this is a useful note. The paper deserves peer review; a serious referee should ask for a simple thickness or sheet resistance calibration and a few replicate measurements, but the qualitative findings are enough to justify publication after moderate revision.","headline":"A useful empirical first look at MnGa chemical stability with a plausible core finding, though the resistance proxy needs validation before quantitative safe-window claims are taken at face value.","tokens_in":5215,"tokens_out":1459,"would_cite":true,"duration_ms":17883,"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":"Water and common acids dissolve MnGa films within minutes, while acetone, ethanol, and brief developer exposure leave them intact.","keywords":["MnGa","L10-ordered ferromagnet","chemical stability","wet etching","spintronics fabrication","deionized water corrosion","TMAOH developer","oxygen plasma passivation"],"falsifier":"Soak a fresh 30-nm MnGa film in deionized water for five minutes, then measure the remaining film thickness and composition by cross-section electron microscopy and XPS: if the resistance has jumped to 20 MΩ while the film thickness and Mn:Ga ratio are unchanged, the resistance-based measure of attack is wrong. Conversely, finding a corresponding thickness loss would confirm the corrosion picture.","tokens_in":4418,"feed_emoji":"🧲","tokens_out":5668,"duration_ms":55162,"temperature":0.7,"pith_summary":"This paper establishes a chemical-safety map for fabricating devices from L10-ordered MnGa, a ferromagnetic film being developed for spintronics. By soaking 30-nm films in standard processing chemicals and watching the electrical resistance of the film, the authors show that acetone and ethanol are harmless, that TMAOH developer is tolerable only for short times, and that deionized water and common acids (HCl, H3PO4, H2SO4) attack MnGa violently within minutes. They also show that oxygen plasma oxidizes the surface, which passivates it. These results give device makers a concrete list of which cleaning, development, and plasma steps are safe and why water-based rinses and acid wet etching should be avoided.","feed_headline":"Water and acids destroy MnGa films within minutes","feed_subtitle":"Resistance tracking shows acetone and ethanol are safe; water destroys MnGa.","key_machinery":"The load-bearing measurement is the electrical resistance of the MnGa layer used as a quantitative proxy for chemical attack: a patterned strip's resistance rises as the conductive film is consumed, so time-resolved resistance curves convert 'the film is attacked' into a number. The supporting material platform is a 30-nm epitaxial MnGa film on semi-insulating GaAs with a 1.5-nm MgO cap, which keeps the as-grown surface intact until each chemical test. Resistance monitoring is what allows the authors to separate benign solvents (constant R), violent corrosives (R reaching 20 MΩ in minutes), slow attack (TMAOH, factor-of-10 rise only after prolonged soaking), and surface passivation (oxygen plasma oxidation) within a single experiment.","core_discovery":"The central claim is that the chemical compatibility of MnGa with standard microfabrication is governed by a few simple rules: water and acid solutions corrode the film aggressively, while organic solvents do not. Measured on 30-nm L10-MnGa films capped with 1.5 nm MgO, the resistance of a 3 mm by 6 mm strip stays constant in acetone and ethanol, rises to 20 MΩ in deionized water within five minutes, changes negligibly in 75% TMAOH developer for the first three minutes but climbs by a factor of ten after extended soaking, and responds to 200 W oxygen plasma by surface oxidation that passivates the film. The paper concludes that acids should be avoided whenever possible, water-based steps reduced to quick rinses, short TMAOH development is safe, and argon-ion milling is the reliable patterning route.","pith_inferences":["The same resistance-monitoring protocol could screen other easily corroded magnetic alloys, such as MnAl or MnBi, against standard fab chemicals before device integration.","Because the paper's own lithography flow already includes a ten-second deionized-water rinse, the violent water reaction implies even short rinses may etch edges; minimizing rinse time or switching to an alcohol rinse is a testable extension.","Oxygen-plasma passivation may alter magnetic properties, not just resistance; measuring anisotropy or coercivity before and after plasma exposure would show whether the passivating oxide is magnetically benign.","The 20 MΩ saturation in water suggests the corrosion product is insulating and may spread through grain boundaries; electron microscopy of the corroded film could identify the attack path and inform protective barrier designs."],"forward_implications":["MnGa device fabrication should not use acid wet etching: HCl, H3PO4, and H2SO4 solutions undercut the resist and remove supposedly protected patterns before unprotected regions are gone.","Cleaning and lift-off with acetone or ethanol are safe; these solvents do not change the film resistance at all.","Photoresist development in TMAOH is acceptable if kept under about one minute; extended soaking raises the film resistance tenfold and should be avoided.","Deionized-water rinses should be replaced by quick dips followed by nitrogen drying; a five-minute water soak can drive the resistance to 20 MΩ.","Oxygen plasma can be used to passivate the MnGa surface by oxidation, but the effect is substantial for ultrathin films and negligible for thick ones, so film thickness must be considered."],"supporting_citations":[{"why":"Supplies the epitaxial growth method and the L10-MnGa film platform used for all stability tests.","marker":"[4]"},{"why":"Describes the MgO capping scheme that keeps the MnGa surface intact before each chemical exposure.","marker":"[5]"},{"why":"Establishes that a thin MgO layer protects MnGa from atmospheric oxidation, justifying the sample handling protocol.","marker":"[18]"}],"fun_headline_variants":["MnGa films dissolve in water but survive organic solvents","Acids and water corrode MnGa; acetone and ethanol safe","MnGa fabrication: avoid water, acids; use argon milling","Oxygen plasma passivates MnGa; water destroys it quickly","MnGa stability: organic solvents fine, aqueous etchants kill"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's quantitative timelines assume that the MnGa film's resistance faithfully measures the degree of chemical attack, without an independent calibration against actual thickness loss, composition change, or contact resistance effects.","fun_headline_variants_meta":{"raw":{"variants":["MnGa films dissolve in water but survive organic solvents","Acids and water corrode MnGa; acetone and ethanol safe","MnGa fabrication: avoid water, acids; use argon milling","Oxygen plasma passivates MnGa; water destroys it quickly","MnGa stability: organic solvents fine, aqueous etchants kill"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000992,"raw_usage":{"total_tokens":4184,"prompt_tokens":905,"completion_tokens":3279,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":521,"completion_tokens_details":{"reasoning_tokens":3193}},"tokens_in":521,"tokens_out":3279,"duration_ms":24708,"temperature":1.0,"reasoning_tokens":3193,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:40:34.803492+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Soak a fresh 30-nm MnGa film in deionized water for five minutes, then measure the remaining film thickness and composition by cross-section electron microscopy and XPS: if the resistance has jumped to 20 MΩ while the film thickness and Mn:Ga ratio are unchanged, the resistance-based measure of attack is wrong. Conversely, finding a corresponding thickness loss would confirm the corrosion picture.","supporting_citations":[{"cited_title":"Multifunctional L10- Mn1.5Ga Films with ultrahigh coercivity, giant perpendicular magnetocrystalline anisotropy and large magnetic energy product","cited_arxiv_id":null,"evidence_quote":"Supplies the epitaxial growth method and the L10-MnGa film platform used for all stability tests."},{"cited_title":"Tailoring magnetism of multifunctional Mn xGa films with 3 giant perpendicular anisotr opy","cited_arxiv_id":null,"evidence_quote":"Describes the MgO capping scheme that keeps the MnGa surface intact before each chemical exposure."},{"cited_title":"Epitaxial MnGa/(Mn,Ga,As)/MnGa trilayers: Growth and magnetic properties","cited_arxiv_id":null,"evidence_quote":"Establishes that a thin MgO layer protects MnGa from atmospheric oxidation, justifying the sample handling protocol."}],"review_version":1}