Pith. sign in

REVIEW 3 major objections 3 minor 33 references

Surface terraces in pure tungsten formed by high-temperature oxidation

T0 review · 3 major / 3 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that residual oxygen, acting preferentially on {100} crystal planes, creates the giant terraces seen on heated polycrystalline tungsten.

desk verdict The terrace observation is real and new, but the preferential-oxidation mechanism rests on an inferred oxygen pressure and a carbon-entangled control, so the paper deserves peer review with a request for a cleaner mechanistic test. read the letter →

arxiv 1908.02036 v1 pith:IPWGGQSI submitted 2019-08-06 cond-mat.mtrl-sci

Hongbing Yu , Suchandrima Das , Junliang Liu , Jason Hess , Felix Hofmann This is my paper · ORCID
classification cond-mat.mtrl-sci
keywords surfaceterracestungstenoxidationorientationdependencehighvacuumannealingoxidesublimationplasma-facingmaterialsEBSD
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Heating pure polycrystalline tungsten to 1500°C in a vacuum furnace with residual oxygen near $10^{-6}$ mbar transforms the surface of near-$\{111\}$ grains into giant terraces whose facets are $\{100\}$ planes, while near-$\{100\}$ grains stay smooth. The paper argues that this restructuring is caused by anisotropic oxidation rather than by surface diffusion: the $\{100\}$ plane oxidises faster, the resulting oxide sublimes above 1000°C, and the un-oxidised $\{100\}$ facets are left behind as tall steps. Because the terraces raise rather than lower the total surface energy, a diffusion-driven mechanism cannot explain them. If the oxidation mechanism is right, texture control becomes a practical lever for slowing oxidation-driven degradation of tungsten in high-temperature, low-oxygen environments such as fusion reactors.

What carries the argument

The carrying mechanism is orientation-dependent oxidation followed by sublimation of the oxide. At about 1500°C with oxygen near $10^{-6}$ mbar, the $\{100\}$ crystal plane of tungsten reacts with oxygen several times faster than $\{111\}$ or $\{110\}$; the volatile oxide sublimes above 1000°C, removing material along $\{100\}$ planes and exposing un-oxidised $\{100\}$ facets as terraces. The furnace-versus-quartz-tube comparison is the key control: when the oxygen supply is cut off, terraces disappear and the surface instead forms W2C and residual oxides.

What would settle it

Heat two polished tungsten crystals with the same near-$\{111\}$ surface orientation at 1500°C for 24 h, one in a vacuum with an oxygen partial pressure near $10^{-6}$ mbar and one in the same vacuum with oxygen gettered below $10^{-9}$ mbar but with identical carbon exposure; if the low-oxygen crystal also develops $\{100\}$-faceted terraces, oxygen is not the controlling agent, and the oxidation mechanism is falsified.

Watch

Extended reading notes

Core claim

The central claim is that preferential oxidation of $\{100\}$ planes controls terrace formation in pure tungsten at high temperature under high vacuum. On near-$\{111\}$ grains, oxygen attacks along $\{100\}$ planes; volatile tungsten oxide leaves the surface, and the remaining material exposes $\{100\}$-faceted steps hundreds of nanometres high. Consistent with this, terraced grains stand 200 nm to 1 μm above neighbouring flat grains, indicating that near-$\{111\}$ surfaces are removed more slowly than near-$\{100\}$ surfaces, and a sample sealed in quartz, which cuts off the oxygen supply, shows no terraces but instead carbide and redeposited oxide features. The paper therefore concludes that the morphology is set by the crystal-orientation dependence of the oxidation rate, not by minimisation of surface energy.

Load-bearing premise

The whole mechanism assumes residual oxygen at roughly $10^{-6}$ mbar is the active agent, while the 35 at% carbon found on the furnace samples is dismissed as post-anneal contamination; if carbon or another impurity actually drives the terracing, the oxidation story collapses.

Editorial extensions

If this is right

  • Vacuum quality alone can restructure tungsten: residual oxygen near $10^{-6}$ mbar over 24 h is enough to etch terraces hundreds of nanometres deep.
  • A tungsten component with a preferred $\{111\}$ texture will oxidise more slowly and stay macroscopically smoother than one exposing $\{100\}$ grains.
  • Surface-energy arguments cannot be used to predict the late-stage morphology of hot tungsten; the reactive environment matters more than equilibrium faceting.
  • Controlling oxygen partial pressure offers a route to deliberately produce $\{100\}$-faceted tungsten surfaces without mechanical processing.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the mechanism holds, oxygen dosing could become a maskless way to pattern tungsten with $\{100\}$ facets for field emitters or catalyst supports, since the facet orientation and step height are set by crystal orientation and oxidation time.
  • A clean controlled experiment with labelled oxygen and carbon-free heating would settle whether the 35 at% carbon on the furnace sample is truly inert; the paper does not rule out a carbon-assisted path.
  • Other refractory bcc metals with volatile oxides, such as molybdenum, should show analogous orientation-dependent terracing under the same vacuum conditions; this is an untested consequence of the proposed mechanism.
  • The terrace height of hundreds of nanometres suggests oxidation proceeds by repeated oxide formation and sublimation cycles; measuring mass loss versus time at fixed oxygen pressure could directly test whether the rate is linear or cyclic.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. The paper reports the formation of large-scale surface terraces on near-{111} grains of pure polycrystalline tungsten after heat treatment at 1500°C in a vacuum furnace at ~1e-5 mbar. Using SEM, EBSD, AFM, and stereo-imaging, the authors show that terrace facets are aligned with {100} planes and the step heights reach hundreds of nanometres, while near-{100} grains remain smooth. They argue against a surface-diffusion mechanism because {100} facets have higher surface energy, and instead hypothesize that preferential oxidation of {100} planes, followed by sublimation of the volatile oxide, drives terrace formation. They support this with a measured mass loss during annealing and a comparison to a sample sealed in a quartz tube with limited oxygen supply, plus height profiles showing that terraced {111} grains stand proud of surrounding smooth grains. The paper also discusses implications for tungsten in fusion reactor environments, suggesting that texture control may mitigate oxidation.

Significance. If the oxidation-driven mechanism is correct, this work identifies a new pathway for surface restructuring of tungsten in high-temperature vacuum environments with residual oxygen, which is directly relevant to plasma-facing components in fusion reactors. The orientation dependence of terrace formation and the mass-loss evidence are consistent with the proposed mechanism, and the paper's suggestion that a (111) texture could be beneficial is an interesting practical implication. The strengths include the combination of crystallographic and topographic characterization, the use of mass loss as a macroscopic oxidation signature, and the explicit comparison of recrystallized and as-rolled samples to rule out recrystallization roughening. However, the central mechanistic claim rests on indirect evidence: the active oxygen partial pressure is inferred rather than measured, and the control experiment with limited oxygen is confounded by high carbon content and W2C formation. These gaps leave room for alternative impurity-driven explanations.

major comments (3)
  1. [Main text, paragraph 'To test this hypothesis of oxidation-driven restructuring' and Fig. 3] The quartz-tube (QT) experiment is not a clean oxygen-limited control. The QT sample shows 50-60 at% surface carbon and a W2C layer, while the furnace sample also has 35 at% surface carbon that is assumed to be post-anneal contamination. This assumption is not established: the carbon could be present during the anneal and could participate in or inhibit terrace formation, for example by stabilizing W2C or altering surface diffusion. The absence of terraces in the QT sample could therefore be caused by carbon or phase changes rather than by oxygen scarcity. A control experiment in which oxygen partial pressure is controlled independently (e.g., with residual gas analysis and a carbon-free furnace environment) is needed to separate oxygen effects from carbon effects.
  2. [Main text, paragraph 'A possible mechanism could be the orientation dependence of oxidation rate'] The oxygen partial pressure in the furnace is not measured; it is inferred to be on the order of 1e-6 mbar from the total pressure of ~1e-5 mbar. The proposed mechanism depends quantitatively on the oxygen flux: it must be high enough to cause measurable oxidation but low enough that the volatile oxide sublimes and does not passivate the surface. Without residual gas analysis or a calibrated oxygen leak, the actual oxygen exposure is unknown, and the argument cannot rule out that carbon, water vapor, or other impurity gases drive the morphology. The mass-loss measurement is consistent with oxidation but does not identify the oxidizing species.
  3. [Main text, paragraphs 'Hence, we hypothesise...' and 'Comparison of the furnace and QT samples...' and Fig. 4] The proposed mechanism predicts preferential etching along {100} planes, yet the observed {100} facets are the ones that would be expected to recede fastest if {100} has the highest oxidation rate. The paper explains this by stating that the un-oxidized {100} facets are exposed after oxide sublimation, but this needs a quantitative or at least a geometric model to be convincing. The height profile in Fig. 4 shows that {111} grains stand proud, which is consistent with a lower overall oxidation rate of {111}, but does not by itself explain why the facets on these grains are {100}. A schematic or a simple kinetic model relating the orientation-dependent oxidation rate to the final facet orientation would strengthen the central claim.
minor comments (3)
  1. [Main text, paragraph 'Our analysis shows...' ] The phrase 'preferentially oxidised along {100} planes' is ambiguous; it would be clearer to state whether the oxidation attacks the bulk along {100} planes or whether the oxide forms preferentially on {100} surface facets.
  2. [Main text, paragraph 'Figure 1'] Subfigure references 'Fig. 1(c-d)' should be written as 'Fig. 1(c)-(d)' for consistency.
  3. [Main text, paragraph 'The presence of terraces...' ] The statement that the total surface energy increase is 'more than 70%' should cite the specific surface-energy values used from refs. [13,14] and state the assumed geometry (e.g., {111} surface replaced by {100} facets at a given projected area).

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the oxidation mechanism rests on external kinetic measurements and is tested by independent morphology and limited-oxygen controls.

full rationale

The derivation chain proceeds from observation to hypothesis, not from a fitted input. The key claim, that preferential oxidation of {100} planes controls terrace formation, is anchored in two prior external measurements of orientation-dependent tungsten oxidation (refs 21 and 22), which report faster oxidation or reaction of {001}-type surfaces at different temperatures. Those rates are not fitted to the terraces observed here, so the facet orientation is not being used as its own explanation. The paper then makes three independent tests: (1) measured mass loss from tungsten oxide sublimation; (2) a limited-oxygen quartz-tube control that produces no terraces but instead W2C and residual oxide; and (3) height profiles showing that terrace-bearing {111} grains stand proud of flat {100} grains, consistent with lower {111} oxidation loss. These tests are not inputs to the hypothesis, so no prediction reduces to the observation by construction. The only self-citation, ref [24], is used for a peripheral explanation of why metastable W2C is retained at room temperature (defect sinks), and it does not support the central terrace mechanism; it is therefore not load-bearing circularity. The unmeasured oxygen partial pressure and carbon contamination in the quartz-tube control are evidentiary weaknesses that could undermine the oxidation mechanism, but they are not circularity: the mechanism's content does not collapse into the data used to support it. Overall, no equation, fitted parameter, or self-citation chain makes the result equivalent to its inputs.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters are fitted: the central claim is qualitative and uses literature surface energies and oxidation rates as inputs. The mechanism rests on several domain assumptions about reaction kinetics, residual oxygen, and the inertness of carbon contamination. No new particles, forces, or mediators are introduced; W2C is an observed phase in the quartz tube sample, not an invented entity.

assumptions (5)
  • domain assumption The surface energy ordering and magnitudes for W from Vitos et al. (ref 13) apply, so (100) has higher energy than (111).
    Used to argue that the observed {100} facets increase total surface energy by more than 70%, ruling out diffusion-driven formation. This is external theoretical data, not measured here.
  • domain assumption Orientation-dependent oxidation kinetics measured at 450-600 C (ref 21) and oxygen reaction probabilities at 2050 C (ref 22) remain valid at 1200-1500 C and ~10^-5 mbar.
    The terrace mechanism relies on extrapolating preferential {100} oxidation to the annealing conditions used in this study.
  • domain assumption WO3 forms and sublimes above 1000 C, exposing the underlying unoxidized tungsten surface.
    This is needed to convert preferential oxidation into faceted terraces; supported by refs 15-20 but not directly measured at the specific pressure and temperature of this experiment.
  • domain assumption Residual oxygen in the vacuum furnace, on the order of 10^-6 mbar, is sufficient to drive the hypothesized oxidation-sublimation cycle.
    The paper infers oxygen leakage from furnace joints but does not directly measure the oxygen partial pressure or the oxidation rate during annealing.
  • ad hoc to paper The high carbon content (35 at%) on furnace samples is post-anneal contamination and does not influence terrace formation.
    No carbon-free control is presented; the quartz tube sample with high carbon gives different morphology, so the role of carbon is not fully separated from oxygen effects.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Surface terraces in pure tungsten formed by high-temperature oxidation." pith.science (2026). https://pith.science/paper/IPWGGQSI

@misc{pith2026190802036,
  author       = {Pith},
  title        = {Pith review of: Surface terraces in pure tungsten formed by high-temperature oxidation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IPWGGQSI}},
  note         = {Machine review of arXiv:1908.02036}
}
read the original abstract

We observe large-scale surface terraces in tungsten oxidised at high temperature and in high vacuum. Their formation is highly dependent on crystal orientation, with only {111} grains showing prominent terraces. Terrace facets are aligned with {100} crystallographic planes, leading to an increase in total surface energy, making a diffusion-driven formation mechanism unlikely. Instead we hypothesize that preferential oxidation of {100} crystal planes controls terrace formation. Grain height profiles after oxidation and the morphology of samples heat treated with limited oxygen supply are consistent with this hypothesis. Our observations have important implications for the use of tungsten in extreme environments.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

  1. [1]

    Brown, P

    T. Brown, P. Pitfield, in: Gus Gunn (Ed.), Crit. Met. Handb., Wiley Online Books, 2013

    work page 2013

  2. [2]

    Williams, C.B

    D.B. Williams, C.B. Carter, Transmission Electron Microscopy, Springer US, 2009

    work page 2009

  3. [3]

    Kecskes, K.C

    L.J. Kecskes, K.C. Cho, R.J. Dowding, B.E. Schuster, R.Z. Valiev, Q. Wei, Mater. Sci. Eng. A 467 (2007) 33–43

    work page 2007

  4. [4]

    Rieth, S.L

    M. Rieth, S.L. Dudarev, S.M. Gonzalez De Vicente, J. Aktaa, T. Ahlgren, S. Antusch, D.E.J. Armstrong, M. Balden, N. Baluc, M.F. Barthe, W.W. Basuki, M. Battabyal, C.S. Becquart, D. Blagoeva, H. Boldyryeva, J. Brinkmann, M. Celino, L. Ciupinski, J.B. Correia, A. De Backer, C. Domain, E. Gaganidze, C. García-Rosales, J. Gibson, M.R. Gilbert, S. Giusepponi, ...

    work page 2013

  5. [5]

    Rieth, J.L

    M. Rieth, J.L. Boutard, S.L. Dudarev, T. Ahlgren, S. Antusch, N. Baluc, M.F. Barthe, C.S. Becquart, L. Ciupinski, J.B. Correia, C. Domain, J. Fikar, E. Fortuna, C.C. Fu, E. Gaganidze, T.L. Galán, C. García-Rosales, B. Gludovatz, H. Greuner, K. Heinola, N. Holstein, N. Juslin, F. Koch, W. Krauss, K.J. Kurzydlowski, J. Linke, C. Linsmeier, N. Luzginova, H. ...

    work page 2011

  6. [6]

    Armstrong, P.D

    D.E.J. Armstrong, P.D. Edmondson, S.G. Roberts, Appl. Phys. Lett. 102 (2013) 251901

    work page 2013

  7. [7]

    Das, D.E.J

    S. Das, D.E.J. Armstrong, Y. Zayachuk, W. Liu, R. Xu, F. Hofmann, Scr. Mater. 146 (2018) 335–339

    work page 2018

  8. [8]

    BOWDEN, K.E

    F.P. BOWDEN, K.E. SINGER, Nature 222 (1969) 977–979

    work page 1969

Show all 33 references
  1. [9]

    R. Butz, B. Krahl-urban, E. Preuss, D. Bruchmann, Phys. Status Solidi 27 (1975) 205–212

    1975

  2. [10]

    The presence of terraces with pre-dominantly {001} facets is surprising

    and [001] axes respectively. The presence of terraces with pre-dominantly {001} facets is surprising. Though reports of the absolute surface energy of tungsten show some variation [13,14], the (100) plane is generally believed to have the highest surface energy followed by the...

    2000

  3. [11]

    Choi, S.K

    D.S. Choi, S.K. Kim, R. Gomer, Surf. Sci. 234 (1990) 262–272

    1990

  4. [12]

    Zuber, Z

    S.M. Zuber, Z. Szczudło, A. Szczepkowicz, Y.. Losovyi, A. Ciszewski, Ultramicroscopy 95 (2003) 165–169

    2003

  5. [13]

    Ferrari, L

    A.M. Ferrari, L. Giordano, N. Rösch, U. Heiz, S. Abbet, A. Sanchez, G. Pacchioni, J. Phys. 10 Chem. B 104 (2000) 10612–10617

    2000

  6. [14]

    Vitos, A

    L. Vitos, A. V Ruban, H.L. Skriver, J. Kollár, Surf. Sci. 411 (1998) 186–202

    1998

  7. [15]

    Wang, E.K

    S.G. Wang, E.K. Tian, C.W. Lung, J. Phys. Chem. Solids 61 (2000) 1295–1300

    2000

  8. [16]

    E. A. Gulbransen, K. F. Andrew, J. Electrochem. Soc. 107 (1960) 619–628

    1960

  9. [17]

    Warren, A

    A. Warren, A. Nylund, I. Olefjord, Int. J. Refract. Met. Hard Mater. 14 (1996) 345–353

    1996

  10. [18]

    Cifuentes, M.A

    S.C. Cifuentes, M.A. Monge, P. Pérez, Corros. Sci. 57 (2012) 114–121

    2012

  11. [19]

    Walsh, J.M

    P.N. Walsh, J.M. Quets, R.A. Graff, J. Chem. Phys. 46 (1967) 1144–1153

    1967

  12. [20]

    Klein, T

    F. Klein, T. Wegener, A. Litnovsky, M. Rasinski, X.Y. Tan, J. Gonzalez-Julian, J. Schmitz, M. Bram, J.W. Coenen, C. Linsmeier, Nucl. Mater. Energy 15 (2018) 226–231

    2018

  13. [21]

    Klein, A

    F. Klein, A. Litnovsky, T. Wegener, X. Tan, J. Gonzalez-Julian, M. Rasinski, J. Schmitz, C. Linsmeier, M. Bram, J.W. Coenen, Fusion Eng. Des. (2019)

    2019

  14. [22]

    Schlueter, M

    K. Schlueter, M. Balden, Int. J. Refract. Met. Hard Mater. 79 (2019) 102–107

    2019

  15. [23]

    Bartlett, J.W

    R.W. Bartlett, J.W. McCamont, J. Electrochem. Soc. 112 (1965) 148–152

    1965

  16. [24]

    Kurlov, A.I

    A.S. Kurlov, A.I. Gusev, Inorg. Mater. 42 (2006) 121–127

    2006

  17. [25]

    H. Yu, Z. Yao, M.R. Daymond, J. Nucl. Mater. 493 (2017) 84–95

    2017

  18. [26]

    Doshi, H

    B. Doshi, H. Xie, C. Zhou, R. Sidibomma, M. Meekins, C. Sborchia, K. Ioki, S. Tyge, A.K. Bhardwaj, G.K. Gupta, in: 2013 IEEE 25th Symp. Fusion Eng., 2013, pp. 1–6

    2013

  19. [27]

    Rappé, Rev

    G.H. Rappé, Rev. Phys. Appl. 12 (1977) 1735–1741

    1977

  20. [28]

    Wegener, F

    T. Wegener, F. Klein, A. Litnovsky, M. Rasinski, J. Brinkmann, F. Koch, C. Linsmeier, Nucl. Mater. Energy 9 (2016) 394–398

    2016

  21. [29]

    Litnovsky, T

    A. Litnovsky, T. Wegener, F. Klein, X. Tan, J. Schmitz, C. Linsmeier, M. Rasinski, A. Kreter, M. Bram, J. Gonzalez-Julian, J. Coenen, Phys. Scr. (2017) 14012

    2017

  22. [30]

    F. Koch, S. Köppl, H. Bolt, Self Passivating W-Based Alloys as Plasma-Facing Material, 2009

    2009

  23. [31]

    Voitsenya, M

    V.S. Voitsenya, M. Balden, A.I. Belyaeva, V.K. Alimov, B. Tyburska-Püschel, A.A. Galuza, A.А. Kasilov, I. V Kolenov, V.G. Konovalov, O.O. Skoryk, S.I. Solodovchenko, J. Nucl. Mater. 434 (2013) 375–381

    2013

  24. [32]

    Parish, H

    C.M. Parish, H. Hijazi, H.M. Meyer, F.W. Meyer, Acta Mater. 62 (2014) 173–181. 11 Supplementary material Fig. S1 Surface morphology of two grains that deviate from {111} surface orientation. The surface orientation in the left grain deviated ~ 30° from the {111} plane, while t...

    2014

  25. [33]

    axes respectively. 12 Fig. S2 The correlation between the surface morphology and the grain orientation from randomly selected regions. Blue grains which have {111} surface orientation show terraces structure on the surface, while green and red grains which are {110} and {001} ...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.