A stable and strongly ferromagnetic Fe₁₇O₁₀ cluster with an accordion-like structure

The study addresses how sub-nanometer iron oxide clusters accommodate spin and structure to yield stable, strongly magnetic species. Iron oxide nanoparticles are important in biomedicine, imaging, microwave absorption, and magneto-optical applications. Moving to sub-nano clusters can tune magnetism and reactivity, with implications for high-density storage and spintronics. Prior work shows many iron oxides favor antiferromagnetism or ferrimagnetism, as Fe atoms are often separated by oxygen without direct Fe–Fe bonds, making strong ferromagnetism and high stability seemingly incompatible. Magic-number transition-metal oxide clusters (e.g., Fe₁₃O₈, Fe₁₂O₁₂) have been identified, but ferromagnetic states are usually less stable than antiferromagnetic ones. Recent mass spectrometry found cubic M₁₃O₈ clusters (M = Fe, Co, Ni) in high-spin states, hinting at the importance of spin accommodation. The authors developed a method to prepare pure anionic Feₙ clusters (n=7–31) and probed their gas-phase reactivity with O₂, discovering a particularly stable and strongly ferromagnetic Fe₁₇O₁₀ cluster with an accordion-like structure.

The paper references extensive prior studies on iron oxides and cluster reactivity: FeO, Fe₂O₃, and Fe₃O₄ show varied magnetic properties (antiferromagnetism and ferrimagnetism). Numerous stoichiometries (e.g., (FeO)ₓ, FeₘOₘ₊₁.₂, (Fe₂O₃)ₓ, and (Fe₂O₃)ₓFeOₓ⁺) have been explored, revealing that Fe atoms are commonly oxygen-separated, favoring antiferromagnetism. Magic-number clusters (Fe₁₀ rings, Fe₁₃O₈ cubes, Fe₁₂O₁₂ cages) have been studied, often stabilizing antiferromagnetic configurations. Size-dependent reactivity has been observed: larger Feₙ clusters preferentially form dioxides upon oxidation and can show increased reactivity beyond certain sizes. Recent work identified neutral Co₁₃O₈ dominance in O₂ reactions (metalloxocubes) and prominent Ni₁₃O₈⁺ and Fe₁₃O₈⁺, with high-spin preferences in cubic M₁₃O₈⁺ (M = Fe, Co, Ni, Rh). Challenges in monitoring anionic iron cluster reactions were noted due to strong magnetic effects. This body of work frames the current investigation into a stable, strongly ferromagnetic iron oxide cluster.

Experimental: A home-built reflection time-of-flight mass spectrometer (Re-TOFMS) with a mini flow tube reactor (inner diameter ≈ 6 mm, length ≈ 60 mm) and a laser vaporization (LaVa) source was used. Anionic Feₙ clusters were generated by ablating a rotating, clean iron disk (99.95%) with a pulsed 532 nm Nd:YAG laser at 10 Hz. Clusters were cooled via supersonic expansion through a nozzle (diameter 1.35 mm, length 35 mm) using pulsed He carrier gas (99.999%, 10.0 atm) delivered by a pulsed valve (Series 9, General Valve). For reactions, 10% O₂ in He (1.0 atm supply) was introduced into the flow tube at room temperature; the internal tube pressure was maintained at 35 Pa. Oxygen doses were controlled via pulse valve on-time (e.g., 185 µs vs 200 µs). Mass spectra were collected for Feₙ⁻ (n=7–31) in the absence and presence of O₂ to monitor product formation (FeₙOₓ⁻). Theoretical: Global structure searches for FeₙOₘ clusters employed the ab initio evolutionary algorithm USPEX interfaced with VASP. Given strong electron correlation and likely high-spin states, GGA+U was used to determine ground-state structures and spin states. A thermodynamic convex-hull analysis compared relative formation enthalpies per atom across FeₙOₘ compositions (referenced to Feₙ and (Fe₂O₃)ₘ). Ab initio molecular dynamics (AIMD) simulations probed thermal stability of Fe₁₇O₁₀ at 300 K and 800 K for 3250 fs with 1 fs time steps, tracking representative Fe–O distances. Reaction pathways and energetics were modeled with Gaussian 09 at the BPW91/6-311G(d) level for Fe and O. Geometry optimizations considered multiple spin states; vibrational frequency calculations provided zero-point corrections. Transition states were confirmed by single imaginary frequency and intrinsic reaction coordinate (IRC) connections. Electronic structure analyses included natural population analysis (NPA) for charge distribution, total and projected density of states (DOS), nucleus-independent chemical shift (NICS) calculations [NICS(0), NICS(1)], and electron localization function (ELF) visualizations.

• Mass spectrometry revealed size-dependent O₂ reactivity: bare Feₙ⁻ clusters (n=7–31) exhibited a Gaussian abundance centered at n=16. Upon low O₂ dose, oxidation products FeₙO₂⁻ formed predominately for n=15–27 with a Gaussian distribution centered at n=20. At higher O₂ dose, oxygen-rich products appeared; notably, Fe₁₇O₁₀ dominated the mass spectrum with prominent abundance, indicating exceptional inertness/stability.
• Reaction schemes considered stepwise oxidation and cluster aggregation: Feₙ⁻ + xO₂ → FeₙO₂ₓ⁻ → … → FeₙOₓ⁻; and FeₙOₓ⁻ + Feₘ⁻ + He → Feₙ₊ₘOₓ⁻ + He.
• Global minima search (USPEX+VASP, GGA+U) identified Fe₁₇O₁₀ ground state with C₂v symmetry: 10 µ₃-O atoms cap 10 hollow sites on an Fe₁₇ framework. The structure can be viewed as fusion of two Fe₁₃O₈ cubic units with an overlapping Fe₃O₆ fragment, resembling an accordion-like form with two low-energy resonating isomers.
• Magnetic properties: The Fe₁₇O₁₀ ground state is strongly ferromagnetic with a total magnetic moment of 56 µB (55 unpaired electrons). A resonant isomer lies only 0.07 eV higher, with an interconversion barrier of 0.17 eV. By comparison, cubic Fe₁₃O₈ optimized with the same method is ferrimagnetic (10 µB).
• Thermodynamic stability: Convex-hull analysis placed Fe₁₇O₁₀ and Fe₁₃O₈ at the lowest positions, indicating prominent thermodynamic stability among surveyed Fe–O compositions. AIMD showed Fe₁₇O₁₀ maintains structural integrity up to 800 K over 3250 fs at 1 fs steps.
• Reaction mechanism insights: DFT pathways show that O₂ approaching Fe₁₇ (side-on or end-on) dissociates spontaneously, forming two µ₃-O atoms on adjacent hollow sites. Similarly, Fe₁₇O₈ + O₂ yields Fe₁₇O₁₀ via spontaneous O–O bond cleavage. In contrast, O₂ on Fe₁₃ does not dissociate spontaneously; end-on superoxo adsorption may convert to peroxo and then dissociate over a small barrier.
• Stepwise O₂ addition thermodynamics for anionic clusters indicate decreasing driving force after Fe₁₇O₁₀: ΔE (eV) for Fe₁₇O₂⁻ + O₂ → Fe₁₇O₄⁻ is −7.03; Fe₁₇O₄⁻ + O₂ → Fe₁₇O₆⁻ is −7.58; Fe₁₇O₆⁻ + O₂ → Fe₁₇O₈⁻ is −8.40; Fe₁₇O₈⁻ + O₂ → Fe₁₇O₁₀⁻ is −8.45; Fe₁₇O₁₀⁻ + O₂ → Fe₁₇O₁₂⁻ is −2.23. The much smaller energy gain beyond Fe₁₇O₁₀ rationalizes kinetic trapping and the observed dominance of Fe₁₇O₁₀.
• Electronic structure: NPA shows similar negative charge on µ₃-O atoms for Fe₁₇O₁₀ and Fe₁₃O₈; NICS over quasi-square surfaces are comparable. However, Fe₁₇O₁₀ exhibits strong spin polarization with dominant α-spin bands over −2 to 7.5 eV and β-spin bands over 0 to 5 eV, yielding an enlarged α-HOMO–LUMO gap. ELF analyses highlight β-aromaticity and localized electron-pair regions consistent with strong ferromagnetism coexisting with high stability.

The findings resolve an apparent contradiction in iron oxide clusters by demonstrating that high thermodynamic and thermal stability can coexist with strong ferromagnetism in a sub-nanometer cluster. The size-dependent reactivity explains why only larger Feₙ (n ≥ 15) clusters undergo efficient O₂ uptake; increased cross-section and vibrational density of states aid energy dissipation and structural relaxation during oxidation. The unique Fe₁₇O₁₀ structure—effectively a fusion of two stable Fe₁₃O₈-like cubes—accommodates a very high spin state (56 µB) with small isomerization barriers, consistent with the observed mass spectral prominence and kinetic trapping at Fe₁₇O₁₀. Thermodynamic convex-hull placement and AIMD stability up to 800 K support robustness. Reaction pathway analyses show facile O–O bond scission on Fe₁₇ (vs Fe₁₃), rationalizing why Fe₁₇O₁₀, not Fe₁₃O₈, dominates under the experimental oxidation conditions. The strong spin polarization and electronic structure (DOS, ELF) underpin the ferromagnetism, suggesting Fe₁₇O₁₀ as a promising building block for spintronic and high-density magnetic storage applications.

Using a customized mass spectrometry platform, the study identifies Fe₁₇O₁₀ as a particularly stable and strongly ferromagnetic iron oxide cluster that dominates oxygenation of larger anionic Feₙ clusters. Global minima searches (USPEX+VASP, GGA+U) determine an accordion-like C₂v ground-state structure with 10 µ₃-O caps and a 56 µB magnetic moment. Convex-hull thermodynamics, AIMD stability up to 800 K, and DFT reaction pathways explain its formation, kinetic persistence, and coexistence of strong ferromagnetism with high stability. The work highlights Fe₁₇O₁₀ as a sub-nano magnetic motif with potential utility in high-density information storage and spintronics. Future research could explore experimental magnetic characterization, substrate or ligand effects, charge-state dependencies, and extension to other transition-metal oxide clusters to generalize the design principles uncovered here.

• Structural and magnetic properties are inferred from DFT (GGA+U) and electronic-structure analyses; direct experimental structural or magnetic measurements (e.g., spectroscopy, magnetometry) were not reported.
• Reaction energetics for O₂ addition were computed for anionic clusters at the BPW91/6-311G(d) level; results may be method-dependent.
• AIMD simulations probed limited timescales (3250 fs) and specific temperatures (300 K and 800 K), which may not capture long-term dynamics or all decomposition pathways.
• Gas-phase conditions (10% O₂ in He, low pressures) and cluster charge states may limit generalizability to condensed phases or supported clusters.
• Although DFT suggests other small clusters (e.g., Fe₁₁) allow O–O dissociation, corresponding products were not observed experimentally, indicating kinetic/experimental constraints not fully captured by theory.