{"id":"c63d6108-2b9f-4e4f-af19-55282b43ee24","arxiv_id":"2506.12941","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":6,"one_line_summary":"Activator solutions for geopolymers are most stable when used warm within about 1 to 1.5 minutes after mixing water, NaOH and sodium silicate in that order, and cooling can push them into an unstable, gelled state.","lead":"This paper models when geopolymer activator solutions, made from water, sodium hydroxide and sodium silicate, stay stable or turn into unusable gels, and finds they should be used within minutes while still warm. A practical rule is to mix water first, then sodium hydroxide, then silicate, and to avoid letting the mixture cool overnight.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation 5's linear temperature translation of the Vail isotherm—using only pure NaOH(aq) and quartz solubility derivatives—is the load-bearing assumption; without a mixed-system check, the Table 5 stability windows are quantitatively unsupported.","rationale":"I reviewed the central argument: the paper aims to quantify a time–temperature stability window for geopolymer activator solutions, with Table 5 and Figure 7 as the headline quantitative output. The weakest link is the construction of the 3D solubility surface: Eq. 5 translates the 25 °C Vail isotherm using only pure-component solubility derivatives, with no mixed-system correction. This is exactly the assumption the reader identified. I considered other possible objections—the spliced speciation model discontinuity at SiO2/Na2O = 2, the absence of code or raw data, and the reliance on prior cooling-model output—but none is more load-bearing than the temperature extrapolation: even a perfect cooling model would inherit error from the solubility surface. The qualitative claims (warm is better, cooling risks precipitation, sequence matters) are backed by direct observations and prior work, so the concern does not justify rejection. It does justify keeping the paper conditional: the quantitative windows need an independent test or clearly stated uncertainty bounds. Since my read matches the reader's verdict, no adjustment is needed.","tokens_in":17971,"tokens_out":4486,"duration_ms":50716,"concrete_test":"Digitize or obtain at least one independent ternary solubility boundary at a temperature other than 25 °C (e.g., Vail's 0 °C or 50 °C data, or modern Na2O–SiO2–H2O isotherms) and refit the US/MS surfaces with a temperature-dependent correction rather than Eq. 5's pure-component translation. Then recompute TλUS and tλUS for solutions A, B, and C, and check whether E remains above the US surface at 25 °C. If a corrected surface moves any crossing by more than about 5 °C, or makes E unstable at ambient, the Table 5 windows and the 'E remains stable' subclaim lose quantitative support; if not, the extrapolation is at least not grossly wrong.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 2.4, the temperature dependence of both the metastable and unstable solubility boundaries is introduced through Eq. 5: C_X(C_SiO2; T_Soln) = C_X(C_SiO2; T_ref) + (∂C_X/∂T)|_{T_ref} (T_Soln − T_ref). Appendix B shows that the temperature derivatives feeding this expression are pure-phase solubilities: B.2/B.3 are the NaOH(aq) saturation curve and its derivative, and B.4/B.5 are the quartz solubility curve and its derivative. No term accounts for mixed Na2O–SiO2–H2O interactions or hydrated sodium silicate solids, even though the boundary being shifted is the manually digitized 25 °C ternary contour from Vail. The quantitative milestones in Table 5—e.g., solution C crossing the unstable surface at 26.7 °C after 7.3 h, and solution E remaining stable at ambient—are therefore the output of an unverified rigid translation of that contour. Even if the cooling model from Skane et al. is accepted, the TλUS values are determined by the intersection with this translated surface. The qualitative conclusions are not weakened: direct observation of A and C precipitating and E remaining translucent supports the ranking. The load-bearing concern is specifically that the predicted window widths, and especially the claim that E stays stable while A and C do not, depend on an untested extrapolation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a combined experimental and modelling study of sodium silicate-based geopolymer activator solutions. Using quantitative 29Si NMR speciation, a thermodynamic cooling model from prior work, and solubility boundaries digitized from a 1952 ternary isotherm, the authors construct 3D solubility hypersurfaces and define time–temperature stability windows for eight activator formulations. They report that activators are ready within about 1–1.5 minutes after mixing, that cooling from the initial stabilisation temperature can push concentrated solutions into an unstable region where precipitation is practically irreversible, and that the preferred feedstock addition sequence is water → NaOH → sodium silicate. The central quantitative claims are the predicted instability temperatures and times in Table 5 (e.g., solution C becomes unstable at 26.7 °C after 7.3 h), while direct observations confirm qualitatively that solutions A and C precipitate upon cooling whereas solution E remains translucent.","tokens_in":18281,"tokens_out":3729,"duration_ms":41936,"significance":"If the quantitative model is reliable, the paper makes a useful practical contribution: it would provide concrete, falsifiable stability windows for activator solutions and challenge the common but imprecise practice of allowing 24-hour 'equilibration' periods. The experimental observations of A/C precipitation and E remaining stable give independent qualitative support for the temperature-driven instability ranking, and the proposed sequence (water → NaOH → silicate) is plausible and practically actionable. The machine-readable (though not code-reproduced) model, the use of established speciation data, and the explicit presentation of fitted polynomial coefficients in Appendix B are strengths. However, the quantitative stability windows, which are the main novelty, rest on an untested temperature extrapolation of the solubility boundaries, and the claimed experimental validation is only qualitative. The paper's significance therefore depends on whether that extrapolation can be justified or the claims appropriately softened.","major_comments":[{"comment":"The temperature dependence of both the metastable and unstable solubility boundaries is implemented by a rigid linear translation of the 25 °C Vail isotherm using partial derivatives taken from pure-component NaOH(aq) and quartz solubility curves (Eqs. B.2–B.5). No correction is included for the concentrated mixed Na2O–SiO2–H2O system, hydrated sodium silicate solids, or the fact that the boundaries being shifted were manually digitized from a ternary contour. Since the quantitative TλUS and tλUS values in Table 5 are obtained by intersecting the cooling trajectory with this translated surface, the numerical windows—including the distinction between C becoming unstable and E remaining stable—are unsupported without either mixed-system solubility data at additional temperatures, a sensitivity analysis with respect to the derivative terms, or an explicit justification for why the pure-component derivatives dominate.","section":"Section 2.4, Eq. (5), Appendix B"},{"comment":"The paper states that the model is 'experimentally validated', but the validation is qualitative: the observations that A and C precipitate and E remains translucent confirm that A and C cross the unstable surface at some temperature during cooling, but they do not confirm the specific predicted values (e.g., C at 26.7 °C after 7.3 h). In addition, the tλUS values are extrapolated from the cooling model of Skane et al. [16], which is itself not re-validated in this paper. The authors should either provide quantitative validation (e.g., controlled cooling experiments that measure the onset of precipitation at known solution temperatures) or revise the validation claim and present the Table 5 values as model predictions that await direct test.","section":"Section 3.2.2, Table 5, Highlights"},{"comment":"The unstable boundary is a two-segment linear regression and the metastable boundary a hexic polynomial fitted to a manually drawn 1952 isotherm, with no uncertainty quantification beyond the R² for the hexic fit. The y-intercept of the unstable boundary is noted to deviate from the literature NaOH solubility (24.13 vs 25 mol/L), yet the consequences of this and of digitization error for λX, TλUS, and tλUS are not examined. A perturbation or sensitivity analysis of Table 5 with respect to boundary coefficients would clarify how robust the central claim—that A and C become unstable while E does not—actually is.","section":"Section 2.4, Eq. (3)"}],"minor_comments":[{"comment":"Equation 7 contains an apparent typo, '(71)', in the last line; it should be '(7)'.","section":"Eq. (7)"},{"comment":"The stability conditions in Table 3 are difficult to parse; for example, the metastable row lists 'C_SiO2 ≤ 0' which seems inconsistent with the surrounding text. Please reformat the conditions clearly.","section":"Table 3"},{"comment":"There are minor typographical errors: 'stabile' in the abstract, 'Akali Activator' in the keywords, and 'Jounral' in reference [17].","section":"Abstract and Keywords"},{"comment":"The combined speciation model has a discontinuity at SiO2/Na2O = 2. The paper dismisses interpolation as unnecessary, but a brief statement of the magnitude of the jump and its effect on model outputs would help the reader assess the uncertainty for compositions near that ratio, especially since the feedstock itself (SiO2/Na2O = 2.4) is above it.","section":"Section 2.3"}],"recommendation":"major_revision","confidential_remarks":"The self-citation to Skane et al. is appropriate because the present model directly extends that prior cooling model. The core issue is that the quantitative stability windows are built on an untested pure-component temperature extrapolation, and the 'experimentally validated' phrasing overstates the evidence. I would ask the authors to either add direct measurements of precipitation onset at controlled temperatures or substantially temper the quantitative claims and reframe the paper around the qualitatively supported ordering and process recommendations."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi Colleague,\n\nHere’s my take on Skane et al. (arXiv:2506.12941). The paper’s real value is qualitative and practical: it gives a clear recommendation for mixing order (water, then NaOH, then silicate), says to use the activator within a few minutes while it’s still warm, and warns that allowing the solution to cool into the “unstable” region produces a gel that is effectively impossible to re-dissolve. That message is supported by direct observation: solutions A and C precipitated, E stayed translucent, and the model’s ranking matches those observations. The authors also deserve credit for an honest limitations section and for building on external datasets rather than on their own fitting alone.\n\nWhat is actually new: the quantified time–temperature stability windows (tλUS and TλUS), the 3D solubility hypersurface visualization, and the explicit sequencing recommendation with stability milestones. The qNMR speciation data for three activator solutions is a useful addition, and the merging of Provis/Harris speciation maps into a hybrid model is a reasonable practical step.\n\nThe soft spot is exactly where the stress-test note points. Equation 5 translates the 25 °C solubility boundaries using temperature derivatives from pure NaOH(aq) and quartz solubility only. The boundary being shifted is the manually digitized ternary contour from Vail (1952), which is a mixed Na2O–SiO2–H2O system. Concentrated alkaline silicate solutions are not ideal mixtures; dissolved silica changes water activity, viscosity, and the identity of the solid that precipitates. No correction is applied for those mixed-system effects. That makes the Table 5 values – for example solution C becoming unstable at 26.7 °C after 7.3 h – quantitatively unsupported. This is load-bearing because those numbers are the paper’s headline. The stress-test note is correct: the qualitative conclusions survive, but the quantitative windows don’t.\n\nTwo minor points: the sequencing evidence is mainly from prior work (Figure 8.2 from [16]), not from new experiments in this paper; and no code or raw data are provided, so the polynomial fits can’t be independently checked.\n\nWho gets value: practitioners and researchers in geopolymer/alkali-activated materials who need a practical guide to activator handling. The paper also works as a teaching example of why extrapolating legacy phase diagrams with pure-component derivatives is risky.\n\nVerdict: this deserves a serious referee. A good reviewer should ask for mixed-system solubility validation or a sensitivity analysis, and for a clearer separation between qualitative and quantitative claims. If the temperature extrapolation is properly tested, the paper becomes a genuinely useful contribution. I’d bring it to a reading group as a cautionary tale about model extrapolation.\n\nBest,\n[Your name]","headline":"A practically useful and honest paper on activator mixing and stability, but the quantitative stability windows rest on an untested temperature extrapolation of a 1952 isotherm; treat Table 5 as provisional.","tokens_in":18805,"tokens_out":3180,"would_cite":true,"duration_ms":34836,"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":"The paper claims that geopolymer activator solutions have a computable time\\u2013temperature window of usable stability: they stabilize about one to one-and-a-half minutes after mixing, and they stay stable only while the cooling solution…","keywords":["geopolymer","activator solution","sodium silicate","solubility modelling","thermochemical stability","29Si NMR","silica speciation","process standardisation"],"falsifier":"Hold a solution with the same composition as Experiment C (final NaOH 8.1 M and a SiO$_2$-to-Na$_2$O molar ratio of 0.9) at 26.7 \\u00b0C for 7.3 hours and check for irreversible gelling; if it remains fluid, the predicted unstable temperature is wrong. A sharper test is to measure the unstable solubility boundary of the ternary H$_2$O\\u2013Na$_2$O\\u2013SiO$_2$ system at, say, 40 \\u00b0C and compare it with the model's linearly shifted surface.","tokens_in":17768,"feed_emoji":"🧪","tokens_out":10104,"duration_ms":100873,"temperature":0.7,"pith_summary":"The paper tries to establish that geopolymer activator solutions\\u2014alkaline mixtures of water, sodium hydroxide, and sodium silicate\\u2014have a computable time\\u2013temperature window of usable stability: they become thermodynamically stable roughly one to one-and-a-half minutes after mixing, and they remain stable only while the cooling solution stays above a temperature-dependent solubility boundary. If true, the common practice of leaving activators to \\u2018equilibrate\\u2019 overnight is unnecessary and can itself trigger irreversible precipitation. The paper also argues that mixing order matters, with water \\u2192 NaOH \\u2192 sodium silicate as the reliable sequence and other orders able to push the same composition into an unstable gel. A sympathetic reader would care because this is a concrete step toward reproducible, standardized activator preparation.","feed_headline":"Geopolymer activators turn to unusable gel on cooling","feed_subtitle":"Model sets the stability window: mix warm, use within minutes, and add water first, then NaOH, then silicate.","key_machinery":"The load-bearing object is the temperature-dependent unstable solubility surface in the H$_2$O\\u2013Na$_2$O\\u2013SiO$_2$ system, built by digitizing the 25 \\u00b0C metastable and unstable boundary contours of the ternary diagram and shifting them linearly in temperature using partial derivatives of NaOH(aq) and quartz solubility (Equation 5). A stability vector $\\lambda_X$ measures how far and in which direction a given solution state must move to hit a boundary; the roots $\\lambda_{MS}(T)=0$ and $\\lambda_{US}(T)=0$ define the temperatures where a cooling solution enters the metastable and unstable regions. Supporting this is a 29Si NMR-based speciation map in $Q^n$ notation ($n$ = number of bridging oxygens on a silicate tetrahedron), which shows that higher temperatures favor depolymerized, more reactive silicate species and that silicate speciation equilibrates within seconds to a couple of minutes.","core_discovery":"The central claim is that activator stability is not set by the SiO$_2$/Na$_2$O ratio alone, but by position in the full H$_2$O\\u2013Na$_2$O\\u2013SiO$_2$ composition\\u2013temperature space. Using 29Si NMR speciation, prior thermodynamic cooling profiles, and solubility contours digitized from a 25 \\u00b0C ternary isotherm, the authors construct metastable and unstable solubility hypersurfaces as functions of SiO$_2$ concentration, NaOH concentration, and temperature. A cooling activator crosses the unstable hypersurface at a temperature $T_{\\lambda\\mathrm{US}}$; for experimental solutions A and C this happens before ambient temperature (30.7 \\u00b0C after 6.0 h and 26.7 \\u00b0C after 7.3 h), while solution E remains metastable at ambient. The model also gives a lower bound of about 1\\u20131.5 minutes for the solution to stabilize after adding sodium silicate. The paper concludes that activators should be used promptly while still warm and prepared in the order water, alkali hydroxide, soluble silicate, because crossing into the unstable region is treated as practically irreversible: reheating does not restore the fluid state.","pith_inferences":["The same hypersurface construction could be re-fit for potassium-silicate or mixed-alkali activators, which the paper does not do; a successful re-fit would test whether the stability-window concept generalizes beyond sodium systems.","The hysteresis claim suggests a direct check: cool an activator briefly into the predicted unstable region, reheat it, and measure whether dissolved silica and fluidity return; the paper observes irreversible gelling but does not report such a reheating assay.","In unheated laboratories the model implies seasonal ambient temperature shifts the effective $t_{\\lambda\\mathrm{US}}$; logging solution temperature rather than elapsed time would be a cheap process control the paper leaves implicit."],"forward_implications":["Activator solutions can be mixed and used within minutes rather than after the \\u226524 h \\u2018equilibration\\u2019 periods common in the literature.","Keeping the solution warm after adding sodium silicate preserves stability, because cooling can cross the unstable solubility surface and trigger practically irreversible gelling.","Mixing order matters independently of composition: water first, then sodium hydroxide, then sodium silicate avoids excursions into the unstable region, while other orders can precipitate unusable gel at identical final composition.","For a given recipe the model defines two numbers, $T_{\\mathrm{Stable}}/t_{\\mathrm{Stable}}$ and $T_{\\lambda\\mathrm{US}}/t_{\\lambda\\mathrm{US}}$, that practitioners can use as a quality-control window.","Dilution, not just the SiO$_2$/Na$_2$O ratio, determines where a solution sits relative to the solubility surfaces, so recipes should be reported by concentration in the ternary H$_2$O\\u2013Na$_2$O\\u2013SiO$_2$ space."],"supporting_citations":[{"why":"Supplies the thermodynamic cooling model, the $T_{\\mathrm{Stable}}$ and $t_{\\mathrm{Stable}}$ values, and the experimental activator compositions used here.","marker":"[16]"},{"why":"Provides the 25 \\u00b0C ternary isotherm whose metastable and unstable boundary contours are digitized for the solubility model.","marker":"[28]"},{"why":"Supplies NaOH(aq) solubility as a function of temperature, used to compute the partial derivative for shifting the unstable boundary.","marker":"[35]"},{"why":"Supplies quartz solubility as a function of temperature, used for the temperature shift of the silica-related boundary.","marker":"[30]"},{"why":"Provides the $Q^n$ speciation model for SiO$_2$/Na$_2$O \\u2264 2 that initializes the speciation surfaces.","marker":"[21]"},{"why":"Provides the speciation model for high-ratio sodium silicate solutions, including colloidal $Q^4$, extending the map to the feedstock composition.","marker":"[32]"},{"why":"Supplies temperature-dependent 29Si speciation trends from which the speciation temperature derivatives are derived.","marker":"[33]"},{"why":"Supports the claim that silicate speciation equilibrates within seconds to a couple of minutes, bounding the lower end of the stability window.","marker":"[38]"}],"fun_headline_variants":["Cooling geopolymer activator? It may be irreversibly gelled","Stability window for activators: warm mix, prompt use","Order in mixing decides: water, then NaOH, then silicate","Model sets activator stability limits in minutes and degrees","Reheating can't fix a cooled geopolymer activator"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The predicted instability temperatures assume the 25 \\u00b0C solubility boundaries can be shifted linearly in temperature using only pure NaOH and quartz solubility data, with no correction for concentrated mixed Na$_2$O\\u2013SiO$_2$\\u2013H$_2$O interactions or hydrated silicate solids.","fun_headline_variants_meta":{"raw":{"variants":["Cooling geopolymer activator? It may be irreversibly gelled","Stability window for activators: warm mix, prompt use","Order in mixing decides: water, then NaOH, then silicate","Model sets activator stability limits in minutes and degrees","Reheating can't fix a cooled geopolymer activator"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000203,"raw_usage":{"total_tokens":1426,"prompt_tokens":1028,"completion_tokens":398,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":644,"completion_tokens_details":{"reasoning_tokens":314}},"tokens_in":644,"tokens_out":398,"duration_ms":4916,"temperature":1.0,"reasoning_tokens":314,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T00:36:54.124908+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Hold a solution with the same composition as Experiment C (final NaOH 8.1 M and a SiO$_2$-to-Na$_2$O molar ratio of 0.9) at 26.7 \\u00b0C for 7.3 hours and check for irreversible gelling; if it remains fluid, the predicted unstable temperature is wrong. A sharper test is to measure the unstable solubility boundary of the ternary H$_2$O\\u2013Na$_2$O\\u2013SiO$_2$ system at, say, 40 \\u00b0C and compare it with the model's linearly shifted surface.","supporting_citations":[{"cited_title":"Comparison of embodied energies of Ordinary Portland Cement with Bayer-derived geopolymer products,","cited_arxiv_id":null,"evidence_quote":"Supplies the thermodynamic cooling model, the $T_{\\mathrm{Stable}}$ and $t_{\\mathrm{Stable}}$ values, and the experimental activator compositions used here."},{"cited_title":"What can vibrational spectroscopy tell about the structure of dissolved sodium silicates?,","cited_arxiv_id":null,"evidence_quote":"Provides the 25 \\u00b0C ternary isotherm whose metastable and unstable boundary contours are digitized for the solubility model."},{"cited_title":"Silicate species of water glass and insights for alkali -activated green cement,","cited_arxiv_id":null,"evidence_quote":"Supplies NaOH(aq) solubility as a function of temperature, used to compute the partial derivative for shifting the unstable boundary."},{"cited_title":"Modeling Speciation in Highly Concentrated Alkaline Silicate Solutions,","cited_arxiv_id":null,"evidence_quote":"Supplies quartz solubility as a function of temperature, used for the temperature shift of the silica-related boundary."},{"cited_title":"Carbon dioxide emissions from traditional and modified concrete. A review,","cited_arxiv_id":null,"evidence_quote":"Provides the $Q^n$ speciation model for SiO$_2$/Na$_2$O \\u2264 2 that initializes the speciation surfaces."},{"cited_title":"MS” and “US","cited_arxiv_id":null,"evidence_quote":"Provides the speciation model for high-ratio sodium silicate solutions, including colloidal $Q^4$, extending the map to the feedstock composition."},{"cited_title":"29Si NMR Spectroscopy of Silicate Solutions. II. On the Dependence of Structure of Silicate Anions in Water Solutions from the Na:Si Ratio,","cited_arxiv_id":null,"evidence_quote":"Supplies temperature-dependent 29Si speciation trends from which the speciation temperature derivatives are derived."},{"cited_title":"The System Na2O-CaO-SiO2-H2O,","cited_arxiv_id":null,"evidence_quote":"Supports the claim that silicate speciation equilibrates within seconds to a couple of minutes, bounding the lower end of the stability window."}],"review_version":1}