Search

The system Fe2O3-SO3-H2O contains the most important minerals of acid mine drainage (AMD), iron oxides, and iron sulfates. For geochemical modeling of the AMD systems, reliable thermodynamic data for these phases are needed. In this work, we have determined thermodynamic data for the most acidic sulfates rhomboclase [(H5O2)Fe(SO4)2·2H2O or (H3O)Fe(SO4)2·3H2O] and the phase (H3O)Fe(SO4)2. The actual compositions of the studied phases are (H3O)1.34Fe(SO4)2.17(H2O)3.06 (molecular mass of 344.919 g/mol) and (H3O)1.34Fe(SO4)2.17 (289.792 g/mol). Structural details for both phases were refined from synchrotron powder X-ray diffraction data. Enthalpies of formation were determined by acid-solution calorimetry. Low-temperature heat capacity was measured for rhomboclase by relaxation calorimetry but a critical analysis of entropies for several oxysalts showed that these data are too high. Entropies for both phases were estimated from a Kopp-rule algorithm. The enthalpies of formation and entropies were combined with previously published temperature-relative humidity brackets to generate an internally consistent thermodynamic data set for rhomboclase: ΔfH° = -3202.03 kJ/mol, S° = 378.7 J/(mol·K); and for (H3O)1.34Fe(SO4)2.17 : ΔfH° = -2276.25 kJ/mol, S° = 253.2 J/(mol·K). Solubility experiments at room temperature and at T = 4 °C agree well with previously reported data in the system Fe2O3-SO3-H2O. An inspection of the extended Pitzer model for Fe3+-SO4 solutions shows that this model reproduces the general topology of the phase diagram, but the position of the calculated solubility curves deviates substantially from the experimental data. Solid-state 2H MAS NMR spectra on deuterated rhomboclase show two isotropic chemical shifts δiso(2H) = of 8 ± 1 and 228 ± 1 ppm, assigned to D5O2 + and Fe-OD2 groups, respectively. Canonical ensemble (NVT) molecular dynamics simulations for (H3O)Fe(SO4)2 at T = 300 K showed that the H3O+ groups maintain their trigonal pyramidal geometry and perform different types of motion.

The local structural environments in a series of natural and synthetic alunite samples [ideally A Al 3 (SO 4 ) 2 (OH) 6 , A = H 3 O + , D 3 O + , Na + , and K + ] have been probed by solid-state 1 H, 2 H, 23 Na, 27 Al, and 39 K NMR spectroscopy. The natural alunite [KAl 3 (SO 4 ) 2 (OH) 6 ] and synthetic hydronium alunite samples contain few structural defects, whereas the synthetic natroalunite and alunite samples have ca. 10% Al vacancies based on 27 Al NMR. A new 27 Al local environment (Al D ) was observed and assigned to Al with one Al vacancy in the first cation sphere. Three different proton environments, Al 2 -OH, Al-OH 2 , and H 3 O + are detected by 1 H and 2 H MAS NMR. The hydronium ion (H 3 O + ) is only observed in hydronium alunite, and is associated with the stoichiometric regions of the sample. It was not detected in 1 H and 2 H NMR spectra of alunite and natroalunite despite K (Na) occupancies of significantly less than 100%, as determined from elemental analysis. Thus, our NMR results suggest that the common assumption, namely that an A vacancy and an Al 3+ vacancy are compensated by adding an H 3 O + and 3 H + (creating 3 Al-OH 2 groups), respectively, is too simplistic. Instead, a significant fraction of the Al 3+ vacancies are compensated for by 4 H + ions, resulting in 4 Al-OH 2 groups per vacancy. This substitution is accompanied by the simultaneous deprotonation of a H 3 O + ion present on the A site. The resultant H 2 O molecule is unnecessary for charge balance, accounting for the A -site deficiency often observed. The presence of Al 3+ and A + vacancies appears closely correlated based on NMR.