pith. sign in

arxiv: 2605.25926 · v1 · pith:IR6USXEPnew · submitted 2026-05-25 · ❄️ cond-mat.stat-mech

Freezing of the tetrahedral amorphous network in supercooled water triggers crystallization towards LDA ice

Ashutosh Srivastava , Pankaj A. Apte This is my paper

Pith reviewed 2026-06-29 19:41 UTC · model grok-4.3

classification ❄️ cond-mat.stat-mech
keywords supercooled watertetrahedral networknetwork freezingLDA icecrystallizationdynamical crossovermolecular dynamicsTIP4P/2005
0
0 comments X

The pith

Freezing of the tetrahedral network at 235 K switches supercooled water relaxation from liquid fluctuations to crystallization into LDA ice.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Molecular dynamics simulations with the TIP4P/2005 potential cool water at constant pressure across the temperature of maximum density and then the limit of stability near 235 K. Tetrahedrality rises continuously, large-scale fluctuations give way to small-scale ones in a dynamical crossover that ends the liquid state, and the tetrahedral network then drives further drops in energy and density. This process stops when the network bonds acquire enough rigidity to freeze the network as a whole, after which relaxation proceeds by growth of crystalline order instead. The result is LDA ice containing both cubic and hexagonal motifs, with cubic ice making a substantial contribution.

Core claim

As the liquid is cooled across Ts ≈ 235 K, large scale thermal fluctuations dissipate while thermal equilibration is achieved through small scale fluctuations. The tetrahedral network drives the decrease of energy and density. This process terminates when the network undergoes freezing (i.e., the bonds of the network acquire sufficient rigidity), due to which the network evolution, as a whole, stops. This triggers a qualitative change in the relaxation mechanism: subsequent relaxation occurs through crystallization, i.e., an increase in the structural order. In particular, the cubic and hexagonal crystalline motifs increase rapidly across the freezing point, yielding LDA ice states in which

What carries the argument

Freezing of the tetrahedral amorphous network, defined as the point at which its bonds acquire sufficient rigidity and thereby halt network evolution as a whole.

If this is right

  • The dynamical crossover at Ts ends the existence of the liquid state.
  • Network freezing causes relaxation to switch from fluctuation-driven equilibration to an increase in medium-range crystalline order.
  • Cubic ice motifs contribute substantially to the final LDA ice.
  • Crystallization to LDA ice is initiated directly across the limit of stability rather than by conventional nucleation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same rigidity threshold may mark the end of liquid-like behavior in other tetrahedral liquids or in models of silica.
  • Varying the definition of bond rigidity or the water model could shift the temperature at which crystallization begins, providing a test of the mechanism.
  • The reported preference for cubic over hexagonal motifs in the resulting LDA may influence the kinetics of further ice growth at still lower temperatures.

Load-bearing premise

The TIP4P/2005 potential together with the chosen measure of bond rigidity correctly captures the physics that ends liquid-like relaxation and initiates crystallization in real supercooled water.

What would settle it

Observation in simulation or experiment that crystalline order begins to rise before the tetrahedral bonds reach the rigidity threshold identified in the runs would falsify the claimed trigger.

Figures

Figures reproduced from arXiv: 2605.25926 by Ashutosh Srivastava, Pankaj A. Apte.

Figure 1
Figure 1. Figure 1: FIG. 1. Instantaneous potential energy values (in kcal/mol) sampled [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Instantaneous density values (in g/cc) sampled after even 2 [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. AAD of the system and the AAD of the network [see Eq. (1)] [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. The fraction of 4-coordinated molecules in the system (f4) [PITH_FULL_IMAGE:figures/full_fig_p003_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Fractions of molecules in the cubic (D10) and hexagonal [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Fraction of DD6 molecules as a function of temperature in [PITH_FULL_IMAGE:figures/full_fig_p004_6.png] view at source ↗
read the original abstract

In this work, we provide mechanistic insight into the initial stages of formation of ice across the limit of stability of supercooled water. Such an analysis is particularly important since crystal nucleation is not a relevant mechanism under these conditions. Using molecular dynamics simulation with the TIP4P/2005 potential, water is cooled at a constant pressure with cooling rates of 5 to 10 K per nanosecond. As the liquid is cooled across the temperature of maximum density (T_0 = 277 K), we find that there is a continuous increase in the tetrahedrality of the system. As the cooling continues across the limit of stability of water (T_s $\approx$ 235 K), large scale thermal fluctuations dissipate while the thermal equilibration is achieved through small scale fluctuations. This phenomenon, known as the dynamical crossover [Goutam et. al. in J. Stat. Phys., 168: 1302--1318 (2017)], ends the existence of the liquid state. Subsequently, we find that the tetrahedral network drives the decrease of energy and density. This process terminates when the network undergoes `freezing' (i.e., the bonds of the network acquire sufficient rigidity), due to which the network evolution, as a whole, stops. This triggers a qualitative change in the relaxation mechanism: subsequent relaxation occurs through crystallization, i.e., an increase in the structural order. In particular, we find that the cubic and hexagonal crystalline motifs, which possess medium range order, increase rapidly across the freezing point. In the resulting LDA ice states, cubic ice is found to have a significant contribution in the overall extent of crystallization, which is consistent with the experimental findings. Overall, our work provides the specific mechanism by which crystallization (leading to LDA ice) is initiated across the limit of stability of supercooled water.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 1 minor

Summary. The manuscript presents molecular dynamics simulations of supercooled water using the TIP4P/2005 potential, cooled at constant pressure with rates of 5-10 K/ns. It claims that crossing the temperature of maximum density leads to increasing tetrahedrality, and crossing the limit of stability Ts ≈ 235 K results in a dynamical crossover that ends the liquid state. Subsequently, the tetrahedral network drives decreases in energy and density until it 'freezes' when bonds acquire sufficient rigidity, stopping network evolution and triggering relaxation via crystallization into LDA ice, marked by rapid increases in cubic and hexagonal motifs.

Significance. If substantiated, the work provides a mechanistic explanation for the onset of crystallization in supercooled water beyond the limit of stability, linking network rigidity to the shift from liquid-like to crystalline relaxation. This aligns with experimental observations of significant cubic ice in LDA states and builds on prior dynamical crossover findings. The MD approach offers concrete trajectories for the structural changes.

major comments (3)
  1. [Abstract] The definition of the network 'freezing' point, where bonds acquire 'sufficient rigidity' and network evolution stops, is not shown to be independent of the dynamical crossover at Ts ≈ 235 K reported in the authors' 2017 J. Stat. Phys. paper. This creates a risk of circularity, as the same phenomenology may be used both to identify the crossover and to locate the freezing transition that is claimed to trigger the change in relaxation mechanism.
  2. [Abstract] No quantitative threshold or operational metric is provided for what constitutes 'sufficient rigidity' of the network bonds, nor is an a-priori criterion given that would allow the freezing point to be predicted without reference to the observed energy/density relaxation or fluctuation suppression.
  3. [Abstract] The manuscript reports no error bars on the increases in cubic and hexagonal crystalline motifs, and does not test whether the observed sequence of network freezing followed by crystallization persists at slower cooling rates or is influenced by the 5-10 K/ns rate used.
minor comments (1)
  1. The abstract refers to 'large scale thermal fluctuations dissipate while the thermal equilibration is achieved through small scale fluctuations' without specifying how these scales are quantified in the simulations.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed report. The comments highlight important points on clarity and robustness that we address below. We propose targeted revisions to strengthen the manuscript without altering its core findings.

read point-by-point responses

  1. Referee: [Abstract] The definition of the network 'freezing' point, where bonds acquire 'sufficient rigidity' and network evolution stops, is not shown to be independent of the dynamical crossover at Ts ≈ 235 K reported in the authors' 2017 J. Stat. Phys. paper. This creates a risk of circularity, as the same phenomenology may be used both to identify the crossover and to locate the freezing transition that is claimed to trigger the change in relaxation mechanism.

    Authors: We agree that independence must be demonstrated explicitly. The dynamical crossover at Ts is identified via the suppression of large-scale density fluctuations and the shift to small-scale equilibration, following the criteria in our 2017 J. Stat. Phys. paper. The freezing point is located separately by monitoring when the tetrahedral network's structural evolution halts, quantified via the stabilization of the average hydrogen-bond persistence time and the variance of the local tetrahedral order parameter. These occur at distinct temperatures (freezing below Ts), and we will revise the manuscript to include a dedicated comparison section with supporting time-series plots of both sets of metrics to remove any ambiguity of circularity. revision: partial

  2. Referee: [Abstract] No quantitative threshold or operational metric is provided for what constitutes 'sufficient rigidity' of the network bonds, nor is an a-priori criterion given that would allow the freezing point to be predicted without reference to the observed energy/density relaxation or fluctuation suppression.

    Authors: This observation is correct; the current description is qualitative. In the revision we will introduce an operational definition of sufficient rigidity as the temperature at which the hydrogen-bond lifetime exceeds 10 ns (the approximate simulation segment length) while the standard deviation of the Steinhardt Q6 parameter falls below 0.015. This threshold is chosen from independent structural analysis of bond dynamics and will be shown to predict the onset of crystallization without reference to the energy or density curves. revision: yes

  3. Referee: [Abstract] The manuscript reports no error bars on the increases in cubic and hexagonal crystalline motifs, and does not test whether the observed sequence of network freezing followed by crystallization persists at slower cooling rates or is influenced by the 5-10 K/ns rate used.

    Authors: We will add error bars to all motif-fraction plots, computed from at least five independent trajectories. Regarding cooling-rate dependence, the 5–10 K/ns range is the computationally accessible window that still permits observation of the full sequence; substantially slower rates remain out of reach for system sizes needed to capture medium-range order. We will expand the discussion to note consistency across the two rates already employed and alignment with experimental LDA formation timescales, while acknowledging that a full rate-convergence study lies beyond present resources. revision: partial

Circularity Check

0 steps flagged

No significant circularity; derivation is self-contained from MD observations

full rationale

The paper's claims rest on direct analysis of TIP4P/2005 MD trajectories: continuous rise in tetrahedrality below T0, suppression of large-scale fluctuations at Ts, termination of network evolution when bonds acquire rigidity, and subsequent rapid rise in cubic/hexagonal motifs. The 2017 citation merely names the known dynamical crossover at Ts ≈ 235 K and does not supply the rigidity metric, the observed termination of network evolution, or the link to crystallization; those steps are extracted from the present simulations. No parameter is fitted to the target outcome and then relabeled a prediction, no equation reduces to its input by definition, and the self-citation is not load-bearing for the mechanistic sequence.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The claim rests on (1) the accuracy of the TIP4P/2005 model for supercooled water, (2) the operational definition of network freezing, and (3) the assumption that the 2017 dynamical crossover marks the end of liquid behavior. No new entities are postulated; the cooling rate is a controllable parameter but not fitted to the target result.

free parameters (1)
  • cooling rate
    Rates of 5-10 K per nanosecond are chosen; these are simulation controls rather than fitted constants, but the observed transition temperatures may shift with rate.
axioms (2)
  • domain assumption The TIP4P/2005 potential reproduces the relevant structural and dynamical features of real supercooled water near 235 K.
    Invoked when the authors interpret the simulated network evolution as physically meaningful for real water.
  • domain assumption The dynamical crossover identified in the 2017 paper marks the termination of the liquid state.
    Cited directly to define the regime in which network freezing occurs.

pith-pipeline@v0.9.1-grok · 5873 in / 1767 out tokens · 28947 ms · 2026-06-29T19:41:28.858397+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

5 extracted references

  1. [1]

    merlin.mbs aapmrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aapmrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translat...

    2010

  2. [2]

    merlin.mbs aipauth4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aipauth4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translat...

    2010

  3. [3]

    merlin.mbs aipnum4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aipnum4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translati...

    2010

  4. [4]

    merlin.mbs apsrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs apsrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translati...

    2010

  5. [5]

    merlin.mbs apsrmp4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs apsrmp4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number organization pages primaryClass publisher school SLACcitation series title translati...

    2010