Controlled polymorphic competition -- a path to tough and hard ceramics

A. Kjell\'en; D.G. Sangiovanni; F. Tasn\'adi; F. Trybel; I.A. Abrikosov; L.J.S. Johnson; M. Od\'en

REVIEW 2 major objections 3 minor 5 references

Controlled polymorphic competition -- a path to tough and hard ceramics

T0 review · 2 major / 3 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Atomic-scale phase switch sets the toughness peak of TiAlN ceramics

desk verdict A credible mechanistic explanation for the toughness maximum of Ti1-xAlxN near x≈0.6, with a real caveat: the x>0.7 decline is read off a 5% transformation threshold rather than unstable crack growth. read the letter →

arxiv 2412.15874 v2 pith:G6EMBEFM submitted 2024-12-20 cond-mat.mtrl-sci

D.G. Sangiovanni , A. Kjellén , F. Trybel , L.J.S. Johnson , M. Odén , F. Tasnádi , I.A. Abrikosov This is my paper · ORCID

classification cond-mat.mtrl-sci

keywords fracturetoughnesstitaniumaluminumnitridetransformationtougheningpolymorphicphasecompetitionatomisticsimulationstress-intensityfactordislocationemissionhardceramiccoatings

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

This paper sets out to show that the fracture toughness and strength of a widely used hard ceramic alloy, cubic (B1) Ti$_{1-x}$Al$_{x}$N, are governed by angstrom-scale plasticity at the crack tip: lattice distortions, a local cubic-to-hexagonal (B1$\to$Bk) phase transformation, and dislocation emission. The authors argue that adding aluminum continuously lowers the energy barrier for that transformation, so the alloy becomes more plastic as $x$ increases. Using stress-intensity-controlled simulations of cracked plates with up to 1.3 million atoms, they predict a toughness and strength maximum near $x=0.6$, with Ti$_{0.4}$Al$_{0.6}$N reaching about 3.2 MPa$\sqrt{\mathrm{m}}$ and 4.3 GPa, followed by a decline for $x>0.7$ when the cubic phase transforms too early into a weaker hexagonal phase. The practical stake is a design rule: tune the energy competition between polymorphs, not just the bond strength, to make hard ceramics tougher.

What carries the argument

The load-bearing mechanism is the tunable energy competition between two named structures: the rocksalt-type cubic phase (B1) and the hexagonal Bk phase, a fivefold-coordinated honeycomb variant of wurtzite. The paper defines two descriptors: $\$\Delta$ E_{B1\to Bk}$, the energy barrier for the tension-driven cubic-to-hexagonal transformation, and $\gamma_f$, the (001) cleavage energy. As $x$ increases, $\$\Delta$ E_{B1\to Bk}$ decreases and so does the unstable stacking-fault energy $\gamma_{\mathrm{usf}}$, so the alloy's response shifts from cleavage to transformation and slip. The argument is carried by the relative position of these two energies: cleavage wins at low $x$, transformation wins too early at high $x$, and near $x\approx0.6$ the gap is just right for slow, localized distortion that absorbs energy before fracture. The Rice criterion for crack-tip dislocation emission is used to confirm that slip is the preferred mechanism for (111) cracks.

What would settle it

Measure the B1$\to$Bk transformation barrier of Ti$_{1-x}$Al$_{x}$N for $x\approx0.85$, with and without applied strain, using a high-level ab initio method and compare it with the potential's prediction; if the barrier is not lower than at $x\approx0.6$, the predicted early-transformation decline fails. A complementary experiment is in-situ straining of Al-rich TiAlN in a transmission electron microscope: observing cleavage without any hexagonal (Bk or B4) domain forming at the crack tip would contradict the proposed mechanism.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that atomic-scale, stress-activated polymorphism controls fracture in Ti$_{1-x}$Al$_{x}$N. For a crack on the (001) plane, the simulations find that TiN fails by pure cleavage, with macroscale toughness $K_{Ic}^{\infty}\approx 1.8$ MPa$\sqrt{\mathrm{m}}$. Adding aluminum makes the lattice around the crack tip pucker, form Bk-like domains, and amorphize locally; these distortions distribute stress and create load-carrying ligaments behind the tip. The predicted toughness and strength rise with aluminum content up to $x\approx0.6$, where $K_{Ic}^{\infty}=3.2$ MPa$\sqrt{\mathrm{m}}$ and $\$sigma_f^{{\infty}}$=4.3$ GPa, exceeding the Griffith purely elastic prediction by a factor of about 2.5. Beyond $x\approx0.7$, the B1 phase transforms into the weaker hexagonal Bk phase at low applied load, and both toughness and strength fall. The same qualitative bell-shaped trend appears for (111) cracks, where plasticity takes the form of dislocation emission and stacking faults, and it matches experimental microcantilever and micropillar data that peak near $x\approx0.6$.

Load-bearing premise

The load-bearing premise is that the classical interatomic force field used for the cracked-plate simulations stays accurate far outside its fitting regime, under the large tensile and shear strains, mixed cubic/hexagonal coordination, and crack-tip disorder that generate the headline toughness values.

Editorial extensions

If this is right

Up to about 60 percent aluminum, each increase in $x$ raises both the fracture toughness and the fracture strength of the cubic alloy, so composition alone is a usable toughening lever for TiAlN coatings.
Beyond about 70 percent aluminum, further alloying is counterproductive in the cubic phase: premature B1$\to$Bk transformation produces a weaker hexagonal phase, giving a distinct cutoff in the useful range.
Because the simulated toughness exceeds the Griffith elastic value by up to a factor of 2.5, a purely elastic fracture theory would miss the dominant mechanism and would wrongly predict that aluminum weakens the material.
The same compositional bell shape appears for (001) and (111) cracks, with the (111) results tying the peak to dislocation emission and stacking-fault formation.
In spinodally decomposing alloys, annealing can create AlN-rich domains that transform more easily, so decomposition period becomes an additional tuning knob for toughness.

Reading between the lines

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

A natural screening rule follows from the paper's two descriptors: compute $\Delta E_{B1\to Bk}$ and $\gamma_f$ from bulk calculations for other hard ceramics and rank candidate alloys by their gap, before expensive fracture simulations are run.
The paper only tests a hexagonal Bk phase that is weaker than the cubic phase; if alloying could stabilize a tougher hexagonal polymorph, the premature-transformation problem at high aluminum content could become an asset instead of a liability.
Because the simulations start from atomically sharp cracks, real coatings with blunt flaws, residual stresses, or mixed-mode loading would most likely shift the optimal composition and the size of the toughness gain; the agreement with experiment suggests the ordering is robust, not that $x\approx0.6$ is a universal constant.
The Bk phase has not yet been detected experimentally in TiAlN; if experiments instead find wurtzite B4 domains at crack tips, the qualitative mechanism would survive but the identity and quantitative energetics of the transformed phase would need revision.

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.

Desk Editor's Note

read the letter

Quick take: the paper gives the first composition-resolved atomistic fracture curves for B1 Ti1-xAlxN and makes a decent case that the toughness maximum near x≈0.6 comes from Å-scale plasticity, specifically tension-driven B1→Bk transformation. That central mechanism is worth taking seriously. The soft spot is the x>0.7 decline: those points are defined by a 5% phase-transformation threshold, not by unstable crack growth, so the “decline” half of the story is more contingent than the rise.

What’s actually new: the KI-controlled methodology is carried over from the authors’ TiN work (Ref. 40) and Curtin’s framework, but the composition sweep from x=0 to 0.95 for both (001) and (111) cracks, the identification of Bk domains forming under tension (a phase not previously associated with crack tips in this system), and the connection to polymorphic energy competition as a design knob are new. The comparison of KIc∞ to Griffith KIcG showing a ratio up to 2.5 is a nice way to quantify the inelastic contribution. The experimental comparison via microcantilever data is honest, including the scatter.

Where it’s soft: the stress-test note is right. For x>0.7, KIc and σf are read at the point where 5% of the plate has transformed to Bk, not at unstable crack extension. The authors say varying Δtip below 5% doesn’t change trends, but that analysis is “not shown” and doesn’t probe larger thresholds. Since the Bk phase is weaker than B1 (they show this), it’s plausible that a constrained transformed domain would trigger instability soon after, but the paper doesn’t demonstrate that. The footnote about KI control degrading as Bk grows is an honest admission but reinforces the concern. The MEAM potential is the other obvious caveat: all quantitative values come from one second-neighbor MEAM fit, and while DFT/AIMD checks confirm qualitative trends in barriers and Bk stabilization, the absolute toughness values are potential-dependent. That’s standard for this kind of work, but worth flagging to any reader who cites the numbers.

Overall: the central maximum near x=0.6 is robust across two crack geometries and matches experiments; the decline for x>0.7 is plausible but rests on a threshold choice. The paper is honest about this. I’d send it to review, and ask the authors to show the threshold sensitivity analysis (including larger Δtip) and release the data or at least the scaling-law forms. It’s a useful paper for anyone working on hard coatings or atomistic fracture, and it gives a concrete mechanistic story that experiments can now probe.

Referee Report

2 major / 3 minor

Summary. The paper uses stress-intensity-controlled atomistic simulations (classical MEAM molecular statics, with DFT/AIMD validation) of cracked B1 Ti1-xAlxN plates with (001) and (111) crack planes to show that Å-scale plasticity—lattice distortions, B1→Bk transformation, and dislocation emission—controls the fracture toughness and fracture strength of these ceramics. The central result is a non-monotonic composition dependence: KIc∞ and σf∞ rise with Al content, reach a maximum near x≈0.6 (e.g., 3.2 MPa√m and 4.3 GPa for Ti0.4Al0.6N(001)), and decline for x>0.7 because of premature B1→Bk transformation. The authors rationalize this behavior through the competition between cleavage energy and polymorphic transformation energy, support it with Rice-criterion analysis and comparison with published microcantilever/pillar measurements, and propose polymorphic competition as a design parameter for hard ceramics.

Significance. If the central claim holds, the paper provides a mechanistic explanation for the experimentally observed toughness and strength maximum in Ti1-xAlxN near x≈0.6 and identifies polymorphic energy competition as a tunable parameter for toughening hard ceramics. The study is commendable for comparing simulations directly with experimental fracture data across a wide compositional range, for considering two crack orientations, for explicit finite-size extrapolation to infinite plate area, and for transparently disclosing limitations. The significance is moderated, however, by the fact that all quantitative KIc∞/σf∞ values come from a single MEAM potential and that the x>0.7 decline is defined by a 5% transformation-extension threshold rather than by unstable crack growth; these two issues are load-bearing for the main claim.

major comments (2)

[Section 2.3 and Section 3.2, footnote 3] The definition of KIc in Section 2.3 sets KIc≡K_I for 0<K_I≤KI5%, where KI5% is a 5% crack-advance or Bk-extension threshold. For x>0.7 the reported decline in KIc∞ and σf∞ (Figure 2a,c) is therefore measured at a transformation-extension threshold, not at the onset of unstable crack growth as defined in Section 2.2. The manuscript defends this cutoff only by 'separate analyses (not shown)' that vary Δtip below 5%, and footnote 3 states that KI control becomes progressively less accurate as Bk grows. Varying the threshold upward is not reported, and the single-phase Bk simulations on (1100)/(1120) cracks do not address the behavior of a constrained Bk domain inside a B1 matrix. If unstable crack growth for x>0.7 occurs at K_I substantially above KI5%, the maximum at x≈0.6 could shift or disappear. Please provide simulations with larger Δtip or an alternative failure criterion (e.g., propagation until unstable crack growth) for Al-rich compositions, or explicitly bound the resulting uncertainty in the reported values.
[Sections 2 and 3.1, Figure 1] All quantitative KIc∞ and σf∞ values are produced by a single second-neighbor MEAM potential (Ref. [38]). Validation against AIMD is limited to small notched models at selected compositions (Section 3.1, Figure 1), and the DFT checks confirm only qualitative trends (barrier decrease with x, Bk stabilization under strain). The central maximum/decline trend, especially the x>0.7 decline, depends on the potential's relative weighting of B1 and Bk energetics near a crack tip. Please provide a quantitative check for at least one Al-rich composition, for example strained B1→Bk transformation barriers from DFT, or results with an independent interatomic potential for one crack geometry, to establish that the predicted decline is not an artifact of the potential.

minor comments (3)

[Figure 3a and Section 3.2] The correlation between the simulated plastic fraction and 1/ΔEB1→Bk is an internal-consistency check, since both quantities are computed with the same MEAM potential. The text should state this explicitly; independent DFT values for ΔEB1→Bk (as in Figure S7) would make the correlation more informative.
[Section 3.4 and Figure 7] The comparison in Figure 7 mixes microcantilever bending and micropillar splitting data and includes dual-phase samples (yellow symbols). The text already cautions against direct comparison, but it would help to mark the testing method and phase composition more clearly in the figure legend or caption.
[Section 2.3] The statement that 'the definitions of σf and KIc are physically meaningful' is supported only by qualitative visualization analyses; please state the quantitative criterion used to verify that the maximum σzz and KI5% coincide with the onset of unstable crack growth or rapid transformation.

Circularity Check

1 steps flagged · score 4.0 of 10

The central maximum near x≈0.6 has independent experimental support, but the x>0.7 toughness decline is partly self-definitional because KIc is defined through a 5% Bk-phase-extension stop criterion, and the paper's defense relies on analyses marked 'not shown'.

self definitional [Sections 2.3 and 3.2 (definition of KI5% and KIc; discussion of the x>0.7 decline)]
"The threshold value KI5% corresponds to a lateral advancement of the crack tip (∆tip), or extension of the secondary Bk phase, equal to 5% of the supercell width. ... This allows us to define the fracture strength σf ≡ σzzmax(K#I) and fracture toughness KIc ≡ K#I for each plate area A. ... The deterioration in mechanical properties in Al-rich B1 Ti1-xAlxN is caused by premature B1→Bk phase transformation, activated under loading."

For x>0.7 the simulations reach KI5% by Bk-phase extension before unstable crack growth. Under the Section 2.3 definition, KIc is then, by construction, the stress intensity at which 5% of the plate has transformed, not the onset of unstable crack growth that Section 2.2 identifies as the physical meaning of KIc. Attributing the low KIc∞ and σf∞ values to 'premature B1→Bk phase transformation' therefore restates the stopping criterion rather than independently demonstrating that transformation causes fracture. The paper's defense ('separate analyses (not shown)') only varies ∆tip below 5% and does not test the alternative that unstable crack growth in the constrained B1+Bk composite occurs at higher stress intensity.

full rationale

The paper is not a case of wholesale circularity. The main quantitative predictions, KIc∞ and σf∞, are produced by stress-intensity-controlled atomistic simulations and then extrapolated with size-scaling relations; they are not fitted to the experimental toughness trend, and the experimental maximum near x≈0.6 (microcantilever and micropillar data) provides external support for the central non-monotonic behavior. The use of the authors' earlier MEAM potential (Ref. [38]) and KI-control method (Ref. [40]) is normal scientific practice: a published interatomic potential is an input model, not an output of this paper, and the paper includes independent DFT/AIMD checks of the transformation-barrier trend and of qualitative fracture behavior. The main definitional concern is the KI5% criterion: when the Bk phase reaches 5% extension before crack growth, the reported 'fracture toughness' is a transformation-onset/early-extension threshold, so the statement that Al-rich alloys are weaker because of premature transformation is partly a restatement of how KIc was defined. The paper acknowledges KI-control degradation in footnote 3 and defends the cutoff only through unshown analyses, which is a transparency weakness rather than a hidden derivation collapse. The plastic-fraction versus 1/ΔE correlation in Figure 3a is likewise an internal consistency check using the same MEAM Hamiltonian, though the DFT-confirmed barrier trend supplies partial independent grounding. Weighing these factors, the central claim retains substantial independent content, but the x>0.7 portion of the trend reduces partly by construction, giving a moderate circularity score of 4.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The quantitative claims rest on three things the reader must take on faith: the transferability of the MEAM potential (Ref. [38]) to crack-tip conditions, the validity of the fitted scaling-law extrapolation to infinite plate area, and the validity of the cubic-based KI field while transformation occurs. No new physical entities are introduced: Bk is a known polymorph, and the helium atoms used to keep notches open in validation models are computational devices, not physical entities.

free parameters (3)

Scaling-law coefficients for KIc(A) and sigma_f(A) extrapolation to infinite area = not reported in main text (Section S1)
KIc∞ and sigma_f∞, the headline macroscale values, are obtained by fitting constitutive scaling laws to finite-area simulation results (Section 2.3). The functional forms and coefficients are not given in the provided text.
Delta_tip = 5% crack-advance or transformation threshold = 5% of supercell width
The operational definition of KIc and sigma_f uses the stress intensity at which crack advance or secondary-phase extension reaches 5% of the plate width (Section 2.3). The paper states results are insensitive to lower thresholds, but only via analyses marked not shown.
MEAM potential parameters for Ti-Al-N (Ref. [38]) = published parameterization of Almyras et al.
All quantitative fracture values inherit the energetics of this classical potential, fitted to Ti-Al-N properties in prior work by the same group. Some trends are cross-checked with DFT and AIMD, but the headline numbers are not independently computed.

assumptions (5)

domain assumption The second-neighbor MEAM potential of Ref. [38] reproduces energetics, plasticity, and fracture of Ti1-xAlxN under large strains at crack tips.
Invoked throughout Sections 2.2 to 3.2; validated only against small notched AIMD models (Section 3.1) and selected DFT quantities.
domain assumption Finite-size simulation results extrapolate to macroscale KIc∞ and sigma_f∞ via the constitutive scaling laws of Section S1.
Section 2.3 states that KIc and sigma_f vary smoothly with plate area and reach asymptotes, with mechanisms asserted to be size-independent based on qualitative inspection.
domain assumption The stress-intensity factor computed from the untransformed cubic elastic response remains a valid loading parameter while B1 to Bk transformation occurs.
Needed for all x>0.7 results; footnote f3 admits the definition of KI becomes progressively less accurate as the hexagonal phase grows.
standard math Griffith (LEFM) and Rice-criterion formulas for KIcG and KIeR apply to disordered Ti1-xAlxN with anisotropic elasticity.
Used for KIcG (Sections 2.3 and S2) and KIeR (Section 3.3, Eq. 6b of Ref. [70]); these are standard linear-elastic fracture mechanics results.
standard math PBE-DFT (and LDA for surface energies) gives reliable relative phase energies and surface energies for Ti1-xAlxN.
Used for validation of MEAM trends (Figures S6 to S9, Table S1); a standard level of theory for this material class.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Controlled polymorphic competition -- a path to tough and hard ceramics." pith.science (2026). https://pith.science/paper/G6EMBEFM

@misc{pith2026241215874,
  author       = {Pith},
  title        = {Pith review of: Controlled polymorphic competition -- a path to tough and hard ceramics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/G6EMBEFM}},
  note         = {Machine review of arXiv:2412.15874}
}

abstract

From nanoscale devices including sensors, electronics, or biocompatible coatings to macroscale structural, automotive or aerospace components, fundamental understanding of plasticity and fracture can guide the realization of materials that ensure safe and durable performance. Identifying the role of atomic-scale plasticity is crucial, especially for applications relying on brittle ceramics. Here, stress-intensity-controlled atomistic simulations of fracture in cubic Ti$_{1-x}$Al$_{x}$N model systems demonstrate how $\overset{\lower.5em\circ}{\mathrm{A}}$-scale plasticity - manifested as lattice distortions, phase transformation, nucleation and emission of dislocations - substantially affects the macroscale fracture toughness (K$_{Ic}$) and fracture strength (${\sigma}$$_{f}$) of brittle ceramics. The extent of plastic deformation in Ti$_{1-x}$Al$_{x}$N increases monotonically with the Al content (x), due to a corresponding decrease in cubic $\rightarrow$ hexagonal polymorph transition energies and unstable stacking fault energies. Overall, plasticity positively affects the mechanical properties, resulting in optimal combinations of strength and toughness for x~0.6. However, for x exceeding ~0.7, the benefits of plasticity diminish. The initial rise followed by a decline in K$_{Ic}$(x) and ${\sigma}$$_{f}$(x) is explained based on the interplay between phase transformation, shear-induced faulting, and tensile cleavage on the easiest fracture plane. The results highlight the impact of atomic-scale plasticity on observable properties and point to strategies for toughening ceramics through control of polymorph competition.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

5 extracted references · 5 canonical work pages

[1]

atomic bridges

Introduction Ceramics are indispensable for technologies that demand a combination of high hardness, heat tolerance, and good chemical inertness. Unfortunately, they are often characterized by a tendency to fracture before yielding. Brittleness is a problem of major concern, especially in safety-critical applications. In contrast to most ceramics, zirconi...

work page arXiv 2000
[30]

Bartosik, C

M. Bartosik, C. Rumeau, R. Hahn, Z.L. Zhang, P.H. Mayrhofer, Fracture toughness and structural evolution in the TiAlN system upon annealing, Scientific Reports 7 (2017) 16476. [31] G. Greczynski, L. Hultman, M. Odén, X-ray photoelectron spectroscopy studies of Ti1–xAlxN (0≤x≤0.83) high-temperature oxidation: The crucial role of Al concentration, Surface &...

work page 2017
[48]

D. Gall, S. Kodambaka, M.A. Wall, I. Petrov, J.E. Greene, Pathways of atomistic processes on TiN(001) and (111) surfaces during film growth: an ab initio study, Journal of Applied Physics 93 (2003) 9086. [49] P. Andric, W.A. Curtin, Atomistic modeling of fracture, Modelling and Simulation in Materials Science and Engineering 27 (2019) 013001. [50] J. Sinc...

work page 2003
[65]

Levämäki, F

H. Levämäki, F. Tasnadi, D.G. Sangiovanni, L.J.S. Johnson, R. Armiento, I.A. Abrikosov, Predicting elastic properties of hard-coating alloys using ab-initio and machine learning methods, npj Computational Materials 8 (2022) 17. [66] D.G. Sangiovanni, D. Edström, L. Hultman, V. Chirita, I. Petrov, J.E. Greene, Dynamics of Ti, N, and TiNx (x = 1–3) admolecu...

work page 2022
[83]

B1 (001)

L. Tan, T. Allen, P. Demkowicz, High temperature interfacial reactions of TIC, ZrC, TiN, and ZrN with palladium, Solid State Ionics 181 (2010) 1156. [84] K. Merkle, D. Smith, Atomic-structure of symmetrical tilt grain-boundaries in NiO, Physical Review Letters 59 (1987) 2887. [85] J. Bean, M. Saito, S. Fukami, H. Sato, S. Ikeda, H. Ohno, Y. Ikuhara, K. Mc...

work page 2010

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

atomic bridges

Introduction Ceramics are indispensable for technologies that demand a combination of high hardness, heat tolerance, and good chemical inertness. Unfortunately, they are often characterized by a tendency to fracture before yielding. Brittleness is a problem of major concern, especially in safety-critical applications. In contrast to most ceramics, zirconi...

work page arXiv 2000

[2] [30]

Bartosik, C

M. Bartosik, C. Rumeau, R. Hahn, Z.L. Zhang, P.H. Mayrhofer, Fracture toughness and structural evolution in the TiAlN system upon annealing, Scientific Reports 7 (2017) 16476. [31] G. Greczynski, L. Hultman, M. Odén, X-ray photoelectron spectroscopy studies of Ti1–xAlxN (0≤x≤0.83) high-temperature oxidation: The crucial role of Al concentration, Surface &...

work page 2017

[3] [48]

D. Gall, S. Kodambaka, M.A. Wall, I. Petrov, J.E. Greene, Pathways of atomistic processes on TiN(001) and (111) surfaces during film growth: an ab initio study, Journal of Applied Physics 93 (2003) 9086. [49] P. Andric, W.A. Curtin, Atomistic modeling of fracture, Modelling and Simulation in Materials Science and Engineering 27 (2019) 013001. [50] J. Sinc...

work page 2003

[4] [65]

Levämäki, F

H. Levämäki, F. Tasnadi, D.G. Sangiovanni, L.J.S. Johnson, R. Armiento, I.A. Abrikosov, Predicting elastic properties of hard-coating alloys using ab-initio and machine learning methods, npj Computational Materials 8 (2022) 17. [66] D.G. Sangiovanni, D. Edström, L. Hultman, V. Chirita, I. Petrov, J.E. Greene, Dynamics of Ti, N, and TiNx (x = 1–3) admolecu...

work page 2022

[5] [83]

B1 (001)

L. Tan, T. Allen, P. Demkowicz, High temperature interfacial reactions of TIC, ZrC, TiN, and ZrN with palladium, Solid State Ionics 181 (2010) 1156. [84] K. Merkle, D. Smith, Atomic-structure of symmetrical tilt grain-boundaries in NiO, Physical Review Letters 59 (1987) 2887. [85] J. Bean, M. Saito, S. Fukami, H. Sato, S. Ikeda, H. Ohno, Y. Ikuhara, K. Mc...

work page 2010