Hidden metallic iron in amorphous silicate dust? Insights from condensation experiments and mid-infrared spectroscopy

Amorphous silicate dust is a major component in the interstellar and circumstellar dust formed in the outflow of asymptotic giant branch (AGB) stars. Although iron depletion is observed in the interstellar medium (ISM), the exact form and fraction of iron in solid remains under debate. In particular, it is unclear whether amorphous silicate dust around AGB stars contains metallic iron. We aimed to provide optical constants of amorphous silicate nanoparticles and examine the effects of metallic iron on their spectral features to better constrain the dust properties by producing amorphous silicate nanoparticles with and without metallic cores. We performed condensation experiments using an induction thermal plasma system to produce dust analogues of the CI chondritic composition in the Mg-Ca-Na-Al-Si-Fe-Ni-O and Mg-Ca-Na-Al-Si-O systems. We measured the absorbance and reflectance of the samples, observed the structure of the products, and determined the optical constants. Two types of amorphous silicate nanoparticles (10-200 nm in diameter) with nearly CI chondritic composition were produced: one contained kamacite (Fe: Ni=0.9: 0.1) cores with a diameter ratio ranging 0-0.87 (average ~0.50), and the other was iron-free homogeneous amorphous silicate. The amorphous silicates of the CI chondritic composition with various sized metallic cores may be prevalent in circumstellar and interstellar dust.

💡 Research Summary

The paper addresses the long‑standing question of how iron is incorporated into the dominant amorphous silicate dust that populates both the interstellar medium (ISM) and the outflows of oxygen‑rich asymptotic giant branch (AGB) stars. While depletion studies show that a large fraction of cosmic iron is missing from the gas phase, the solid‑state carrier—whether it is locked in silicate lattices, exists as separate metallic grains, or is embedded as nanoinclusions—remains uncertain. To resolve this, the authors performed laboratory condensation experiments using an induction thermal plasma (ITP) system, a technique capable of vaporising refractory oxides at temperatures of order 10⁴ K and then quenching the vapor at rates of 10⁴–10⁵ K s⁻¹, thereby mimicking the rapid cooling expected in stellar outflows.

Two experimental runs were designed. The first (CI‑1) employed a CI chondritic composition (Mg‑Ca‑Na‑Al‑Si‑Fe‑Ni‑O) under mildly reducing conditions (oxygen fugacity X_O ≈ 0.93–0.98). This regime favours the condensation of Fe‑Ni metal (kamacite, Fe:Ni ≈ 0.9:0.1) before silicate formation, leading to nanoparticles that consist of a metallic core surrounded by an amorphous silicate mantle. The second run (CI‑2) used the same elemental ratios but omitted oxygen‑bearing species for Fe and Ni, instead adding O₂ to the plasma to create a highly oxidising environment (X_O ≈ 9.35–9.40). Under these conditions, iron remained fully oxidised and the resulting particles were homogeneous, iron‑free amorphous silicates.

Bulk phase analysis (X‑ray diffraction) showed broad halos characteristic of amorphous material in both runs, with distinct kamacite peaks only in CI‑1. Electron probe microanalysis confirmed that the overall bulk composition of the condensates matches the CI chondritic ratios within experimental uncertainty. Transmission electron microscopy combined with energy‑dispersive X‑ray spectroscopy revealed that the CI‑1 particles have diameters between 10 and 200 nm, with core‑to‑total diameter ratios ranging from 0 to 0.87 (average ≈ 0.5). The CI‑2 particles displayed a uniform amorphous structure without any metallic inclusions.

Optical constants were derived from Fourier‑transform infrared (FT‑IR) transmission and reflectance measurements. The authors prepared KBr pellets for transmission and gold‑coated pellets for reflectance to minimise scattering artefacts. By fitting the measured spectra with a Lorentz oscillator model, they extracted complex refractive indices (n + ik) across the mid‑infrared (MIR) range (≈ 2.5–25 µm). The presence of a metallic core subtly shifted the classic Si–O stretching (≈ 10 µm) and Si–O–Si bending (≈ 18 µm) bands, broadening them and moving peak positions by a few tenths of a micron. More strikingly, the near‑infrared (NIR) absorption (2–5 µm) increased by an order of magnitude for core‑shell particles, reflecting free‑electron contributions from the Fe‑Ni metal. Reflectance data showed higher albedo for the core‑shell samples, consistent with enhanced scattering from the conductive core.

Using Mie theory with the newly measured optical constants, the authors demonstrated that metallic inclusions can raise the absorption efficiency (Q_abs) of sub‑micron silicate grains in the NIR by factors of 2–5, while only modestly affecting the MIR band strengths. This has direct implications for radiative‑transfer models of AGB star envelopes: the dust opacity in the NIR, which drives radiation pressure and wind acceleration, could be significantly underestimated if metallic iron cores are ignored. Moreover, the subtle MIR band modifications provide a potential diagnostic for detecting metallic inclusions in astronomical spectra, especially when high‑resolution data (e.g., from JWST/MIRI) are available.

The study also connects its laboratory analogues to the GEMS (glass with embedded metal and sulfides) grains found in chondritic porous interplanetary dust particles. GEMS exhibit similar core‑shell morphologies and IR spectral signatures, supporting the hypothesis that at least a fraction of interstellar silicate dust originates from AGB outflows that produce such composite grains. The authors argue that the estimated 70 % of depleted iron residing in metallic form within the ISM (as suggested by numerical simulations) could be naturally explained by the widespread formation of these core‑shell nanoparticles.

In conclusion, the paper provides the first experimentally derived optical constants for CI‑composition amorphous silicate nanoparticles both with and without embedded Fe‑Ni metallic cores. It shows that metallic iron, when incorporated as nanoinclusions, subtly reshapes MIR silicate features while dramatically enhancing NIR opacity. These results supply a concrete, physically realistic alternative to the historically “astronomical silicate” optical constants, enabling more accurate modeling of dust emission, absorption, and dynamics in circumstellar and interstellar environments. Future work is suggested to explore a broader range of core‑to‑mantle ratios, non‑spherical morphologies, and the inclusion of other transition metals, thereby refining our understanding of dust evolution from stellar sources to the diffuse ISM.