Fe17W3 synthesis-feasibility brief: two nonequilibrium routes, both unproven at the ordering step

Question. Fe17W3 is a computational prediction: tetragonal P-4m2 (#115), 20 atoms/cell, e_above_hull 0.0087 eV/atom by independent replication, every property route-predicted (evidence dossier, forensics). Could a lab actually make it, and how would we know if we did? This brief proposes two fabrication routes, states what each borrows from literature versus what is speculative, and gives the signatures that would settle it. It is a test plan for a prediction, not a recipe: the most likely bulk outcome at this composition is the equilibrium two-phase mixture, and seeing that null result cleanly is itself the check.

Target composition

Fe17W3: 85.0 at% Fe / 15.0 at% W, strictly binary, no interstitials. Predicted structure is tetragonal P-4m2 #115, 20 atoms/cell, density 10.275 g/cm3, layered chemistry (mixed W/Fe planes alternating with pure Fe planes). At 9 meV/atom above the hull it is metastable: slow equilibrium cooling will not produce it. It must be frozen in by nonequilibrium processing, which is what constrains every route below.

What the phase diagram forbids and permits

The assessed Fe-W diagram (see the literature note

Route 1: co-sputtering onto cold substrates, then a low-temperature ordering anneal

• Deposition (literature-supported): co-sputter Fe and W onto water-cooled or LN2-cooled substrates, tuned to 15.0 at% W, staying inside the documented single-bcc window (x < 0.2) and below the amorphous onset near 20 at% W.

• Ordering anneal (speculative): a low-temperature anneal, nominally 300-500 °C, intended to let Fe/W order into the predicted P-4m2 layered arrangement while staying below the regime where the equilibrium alpha + Laves mixture wins. No experiment has ever ordered tetragonal Fe17W3, and the phase-diagram driving force points the other way, toward Fe2W; this anneal is the hypothesis, not the method.

• Combinatorial variant (literature-supported method): a graded-W library spanning roughly 10-20 at% W on one wafer, so the actual single-phase ordering window is measured rather than assumed to sit at exactly 15.0 at%.

Route 2: mechanical alloying plus a short low-temperature anneal

• Supersaturation (literature-supported): mechanically alloy Fe and W powders at 15 at% W; Jartych et al. showed Fe(W) solid-solution formation far beyond equilibrium solubility, with nanocrystalline grain size intrinsic to the process.

• Ordering anneal (speculative): short anneals below about 600 °C to attempt ordering without equilibrating to alpha + Laves. Nanocrystallinity matters here for a second reason: the only known precedent for an unknown W-enriched tetragonal phase in a Fe-bearing system is a metastable nanophase precipitated under nonequilibrium processing in Fe-Co-Cr-W-Ga alloys, unresolved structurally and paramagnetic, which shows Fe-rich Fe-W-containing systems can host such phases when processing freezes them in.

Competing phases (what these routes will actually produce)

1. bcc alpha-Fe(W): the default product of both routes. Ferromagnetic, and nearly pure iron, so its Ms exceeds the candidate's predicted value. XRD, not magnetometry, separates it from Fe17W3.

2. lambda-Fe2W, C14 Laves: the equilibrium partner at 15 at% W; a weak ferrimagnet (experimental low-T Ms ~0.44 T). Its appearance signals the anneal crossed into equilibrium territory.

3. mu-Fe7W6: stable only at much higher W content and 1215-1637 °C; unlikely at 15 at% W, but it shares the diagram with the above and is cheap to check for.

4. Amorphous fraction: expected above ~20 at% W in sputtered films and possible under aggressive milling; a broad halo in XRD.

Checkable signatures

The magnetometry row is the cheapest discriminator and should be run first on any product: a Ms far above the alpha + Laves mixture's phase-weighted value, in a sample whose XRD rules out pure bcc alpha-Fe(W), is the single strongest signal the predicted phase is real. A Ms consistent with an alpha + Laves phase-fraction mixture is the prediction failing cleanly.

Labels, stated plainly

• Sputter-deposition window at 15 at% W: literature-supported (Sumiyama 1991; Lu and Chien 1990; Russell 2025).

• Mechanical-alloying supersaturation at 15 at% W: literature-supported (Jartych 2000).

• Ordering of bcc supersaturation into tetragonal P-4m2 Fe17W3 under any anneal: speculative. No prior report, and equilibrium thermodynamics opposes it. Both routes are vehicles for testing this one step.

• Excluded up front: any melt route (phase diagram), and interstitial-stabilized variants (the program's H1/H4 refutations showed carbon/boron additions fail their own gates in this space).

Every Fe17W3 property cited here is a route prediction with its action receipt in the candidates dataset, not a measurement. If a lab makes this phase, the first number to trust is their magnetometer, not our predictor.

Sources

1. K. Sumiyama, M. Hirata, W. Teshima, "Magnetic and electrical properties of nonequilibrium Fe-W alloys produced by sputter deposition", Jpn. J. Appl. Phys. 30, 2839 (1991). DOI: 10.1143/jjap.30.2839

2. M.-H. Lu, C. L. Chien, "Structural and magnetic properties of Fe-W alloys", J. Appl. Phys. 67, 5787 (1990). DOI: 10.1063/1.345964

3. E. Jartych, J. K. Zurawicz, D. Oleszak, M. Pekala, "Structure and magnetic properties of mechanosynthesized iron-tungsten alloys", J. Magn. Magn. Mater. 218, 247 (2000). DOI: 10.1016/S0304-8853(00)00394-2

4. K. B. Russell, C. A. Kohnke, J. R. Trelewicz, A. M. Hodge, "Investigating phase regimes via combinatorial synthesis: a pathway to tailored materials libraries", Mater. Des. (2025). DOI: 10.1016/j.matdes.2025.113881

5. O. Kubaschewski von Goldbeck, "Fe-W Iron-Tungsten", in Iron - Binary Phase Diagrams, pp. 164-167 (1982). DOI: 10.1007/978-3-662-08024-5_76

6. A. Antoni-Zdziobek, T. Commeau, J.-M. Joubert, "Partial Redetermination of the Fe-W Phase Diagram", Metall. Mater. Trans. A 44, 2996 (2013). DOI: 10.1007/s11661-013-1658-2

7. A. Edstrom, "Magnetocrystalline anisotropy of Laves phase Fe2Ta1-xWx from first principles",

Full phase-diagram sourcing and caveats (including the partially verified 1983 Calphad citation) are in the literature note