Sorption of noble gases by solids, with reference to meteorites. I. Magnetite and carbon

Abstract

To simulate trapping of meteoritic noble gases by solids, 18 samples of Fe 3O 4 were synthesized in a noble gas atmosphere at 350–720 K by the reactions: 3Fe + 4H 2O → Fe 3O 4 + 4H 2 ( Ne, Ar, Kr, Xe) 3Fe + 4CO → Fe 3O 3 + 4C + carbides ( Xe only) Phases were separated by selective solvents (HgCl 2, HCl). Noble gas contents were analyzed by mass spectrometry, or, in runs where 36 d Xe 127 tracer was used, by γ-counting. Surface areas, as measured by the BET method, ranged from 1 to 400 m 2/g. Isotopic fractionations were below the detection limit of 0.5%/m.u. Sorption of Xe on Fe 3O 4 and C obeys Henry's Law between 1 × 10 −8 and 4 × 10 −5 atm, but shows only a slight temperature dependence between 650 and 720 K ( ΔH sol = −4 ± 2 kcal/mole ). The mean distribution coefficient K Xe is 0.28 ± 0.09 cc STP/g atm for Fe 3O 4 and only a factor of 1.2 ± 0.4 greater for C; such similarity for two cogenetic phases was predicted by Lewis et al. (1977). Stepped heating and etching experiments show that 20–50% of the total Xe is physically adsorbed and about 20% is trapped in the solid. The rest is chemisorbed with ΔH s − −13 kcal/mole . The desorption or exchange half-time for the last two components is >10 2 yr at room temperature. Etching experiments showed a possible analogy to “Phase Q” in meteorites. A typical carbon + carbide sample, when etched with HNO 3, lost 47% of its Xe but only 0.9% of its mass, corresponding to a ~0.6 Å layer. Though this etchable, surficial gas component was more thermolabile than Q (release T below 1000°C, compared to 1200–1600°C), another experiment shows that the proportion of chemisorbed Xe increases upon moderate heating (1 hr at 450°C). Apparently adsorbed gases can become “fixed” to the crystal, by processes not involving volume diffusion (recrystallization, chemical reaction, migration to traps, etc.). Such mechanisms may have acted in the solar nebula, to strengthen the binding of adsorbed gases. Adsorbed atmospheric noble gases are present in all samples, and dominate whenever the noble gas partial pressure in the atmosphere is greater than that in the synthesis. Many of the results of Lancet and Anders (1973) seem to have been dominated by such an atmospheric component; others are suspect for other reasons, whereas still others seem reliable. When the doubtful samples of Lancet and Anders are eliminated or corrected, the fractionation pattern—as in our samples—no longer peaks at Ar, but rises monotonically from Ne to Xe. No clear evidence remains for the strong temperature dependence claimed by these authors.