Trace element incorporation into pyrope–grossular solid solutions: an atomistic simulation study

 We have performed atomistic computer simulations on trace element incorporation into the divalent dodecahedral X-sites of pyrope (Py — Mg₃Al₂Si₃O₁₂) – grossular (Gr — Ca₃Al₂Si₃O₁₂) solid solutions. An ionic model and the Mott–Littleton two-region approach to defect energies were used to calculate the energetics of substitution by a range of divalent trace-elements and of charge-balanced substitution by trivalent ions in the static limit. Results are compared with experimental high-temperature, high-pressure garnet-melt trace element partitioning data obtained for the same garnet solid solution to refine our understanding of the factors controlling element partitioning into solid solutions. Defect energies (U _def,f), relaxation (lattice strain) energies (U _rel), and solution energies (U _sol) were derived using two different approaches. One approach assumes the presence of one type of hybrid X-site with properties intermediate between pure Mg and Ca sites, and the other assumes discrete Mg and Ca X-sites, and thus two distinct cation sublattices. The hybrid model is shown to be inadequate, since it averages out local distortions in the garnet structure. The discrete model results suggest trace elements are more soluble in Py₅₀Gy₅₀ than in either end-member compound. Physically this is due to small changes in size of the X-sites and the removal of unfavourable interactions between third nearest neighbours of the same size. Surprisingly, depending on the local order, large trace element cations may substitute for Mg²⁺ and small trace elements for Ca²⁺ in Py₅₀Gr₅₀. These solubilities provide an explanation for the anomalous trace-element partitioning behaviour along the pyrope–grossular join observed experimentally. Received: 27 January 2000 / Accepted: 14 February 2003     

Molecular simulations of interfacial and thermodynamic mixing properties of grossular–andradite garnets

Experimental observations using transmission electron microscopy (TEM) indicate that Fe³⁺-rich grossular–andradite solid solutions with oscillatory zoning tend to occur as separate lamellae of andradite and intermediate compositions (Hirai and Nakazawa 1986; Pollok et?al. 2001). From one lamella to the next, the Fe³⁺ concentration can change significantly within a few nm. In order to understand the Fe³⁺ and Al content of each phase and the thermodynamics, chemistry, structure, and stability at the interfaces, Monte Carlo simulations were performed. According to our calculations, there is an ordered structure with a 1:1 ratio of Al and Fe³⁺ with alternating Al and Fe octahedra along the main cubic crystallographic axes. Even though this ordered grandite is more energetically favorable than a 1:1 mixture of the end members grossular and andradite [by ≈1.6?kJ (mol exchangeable cations)^?1], this structure is stable only at temperatures below ≈500?K. Enthalpies, free energies, configurational and vibrational entropies of mixing, and the long-range order parameter are influenced by the formation of ordered grandite below 500?K. These data also explain why interfaces are stable only between grossular and grandite or between andradite and grandite but not between the end members. The interface energies between the end members and ordered grandite are comparably low [0.16?meV?Å^?2∥(1?0?0), 0.55?meV?Å^?2∥(1?1?0), 0.63?meV?Å^?2∥(1?1?1)] and, therefore, do not hinder the formation of lamellae. Our calculations on the free energies of mixing indicate that there are miscibility gaps between grossular and grandite and between grandite and andradite only below ≈430?K. Since most of these solid solutions are formed at higher temperatures for which we did not find evidence of a miscibility gap, the formation of compositional oscillations is probably due to kinetic hindering of thermodynamically stable complete solid solutions. ?A new methodological aspect is the incorporation of zero-point energies of vibrations and the vibrational entropies into the calculation of the free energy of mixing. In case of the grossular–andradite solid solution, these vibrational effects change the free energy of mixing by only a few percent.     

The crystal chemistry and Fe<Superscript>II</Superscript>–site properties of aluminosilicate garnet solid solutions as revealed by Mössbauer spectroscopy and electronic structure calculations

The three binary garnet solid solutions Fe^II₃Al₂Si₃O₁₂–X^II₃Al₂Si₃O₁₂ (X^II= Mg^II, Mn^II, Ca^II) have been investigated by ⁵⁷Fe Mössbauer spectroscopy at 298 and 77 K and by electronic structure calculations in the local spin density approximation. The spectra yield isomer shifts and quadrupole splittings that are typical for Fe^II in the dodecahedral X-site of 222 point symmetry and are similar for each of the three binaries recorded. Conversely, electronic structure calculations based on the experimental crystal structure of the different end-member garnets exhibit pronounced variations in some of the electronic properties of Fe^II that are not reflected in the spectroscopic data. These results are interpreted as indicating that the different X–O bonds in garnet solid solutions retain to a large degree the intrinsic lengths that they possess in their respective end members, and that the Fe–O bond does not change greatly as a function of composition. This is evidence for the state of alternating bonds and not for the virtual crystal approximation in describing the X–O bond types or lengths in aluminosilicate garnet solid solutions. The observed degree and behavior of the Fe^II doublet asymmetry in the Mössbauer spectra for the three solid solution series do not indicate major variations in the anisotropic recoil-free fraction of Fe^II. Variations in doublet asymmetry are more likely a result of complex next-nearest X-site neighbor interactions and/or some degree of short-range cation ordering, though doublets representing different local X-site cation configurations cannot be resolved or fitted to the experimental spectra.     

Monte Carlo simulation of mixing in Ca3Fe2Ge3O12–Ca4Ge4O12 garnets and implications for the thermodynamic stability of pyrope–majorite solid solution

Static lattice energy calculations, based on empirical pair potentials, were performed for a large set of structures differing in the arrangement of octahedral cations within the garnet 2 × 2 × 2 supercell. The compositions of these structures varied between Ca₃Fe₂Ge₃O₁₂ and Ca₄Ge₄O₁₂. The energies were cluster expanded using pair and quaternary terms. The derived ordering constants were used to constrain Monte Carlo simulations of temperature-dependent mixing properties in the ranges of 1,073–3,673 K and 0–10 GPa. The free energies of mixing were calculated using the method of thermodynamic integration. The calculations predict a wide miscibility gap between Fe-rich (cubic) and Fe-pure (tetragonal) garnets consistent with recent experimental observations of Iezzi et al. (Phys Chem Miner 32:197–207, 2005). It is shown that the miscibility gap arises due to a very strong cation ordering at the Fe-pure composition, driven by the charge difference between Ca²⁺ and Ge⁴⁺ cations. The structural and thermodynamic analogies between Ca–Ge and Mg–Si systems suggest that a similar miscibility gap should exist between pyrope and Mg–Si-majorite.     

Simulation of thermodynamic mixing properties of garnet solid solutions at high temperatures and pressures

We present results of high temperature, high pressure atomistic simulations aimed at determining the thermodynamic mixing properties of key binary garnet solid solutions. Computations cover the pressure range 0–15 GPa and the temperature range 0–2000 K. Through a combination of Monte-Carlo and lattice-dynamics calculations, we derive thermodynamic mixing properties for garnets with compositions along the pyrope–almandine and pyrope–grossular joins, and compare these with existing experimental data. Across the pressure–temperature range considered, simulations show virtually ideal mixing behaviour in garnet on the pyrope–almandine join, while large excess volumes and enthalpies of mixing are predicted for garnet along the pyrope–grossular join. Excess heat capacities and entropies are also examined. These simulations shed additional light on the link between the behaviour at the atomic level and macroscopic thermodynamic properties: we illustrate the importance of certain atomistic Ca–Mg contacts in the pyrope–grossular solid solutions. For simulation techniques of this type to become sufficiently accurate for direct use in geological applications such as geothermobarometry, there is an urgent need for improved experimental determinations of several key quantities, such as the enthalpies of mixing along both joins.     

Direct calculation of thermodynamic properties of the barite/celestite solid solution from molecular principles

 Thermodynamic properties of the barite–celestite solid solution were calculated using molecular principles. Cation–cation (Ba–Ba, Sr–Sr, and Ba–Sr) interaction energies were derived from a number of random and ordered cation distributions which were energy-optimized using force potentials as incorporated in the program package GULP. With these interaction energies, diagrams for the enthalpy and free energy of mixing could be computed for the entire range of the solid solution between the barite and celestite end members and for a number of annealing temperatures. These thermodynamic data show that the solid solution is nonideal. The system has a tendency for Ba²⁺ and Sr²⁺ cations to order onto alternating layers ||(100). However, this ordering scheme is thermodynamically only relevant for annealing temperatures below approximately 500 K and systems that are kinetically inhibited during crystal growth. For sufficiently long annealing times at room temperature, the solid solution tends to exsolve with barite–celestite interfaces ||(100). The cell parameters a and c were calculated to have almost linear behavior for the whole solid solution, suggesting close to ideal behavior according to Vegard's law. In contrast, b tends to deviate positively from linearity, in agreement with experimental values. Received: 6 April 1999 / Revised, accepted: 29 September 1999     

Calculation of thermodynamic properties of hydrated borates by group contribution method

 General equations to correlate and predict the thermodynamic properties of hydrated borates were developed based on the experimental results according to their structural types. The thermodynamic properties (ΔH _f ⁰ and ΔG _f ⁰) of a hydrated borate phase are the sum of the contributions of the cations in aqueous solution, the borate polyanions, and the structural water to the corresponding thermodynamic properties. This method is called the group contribution method, and it is extensively used to calculate the thermodynamic properties of many kinds of inorganic compounds, such as silicates and clay minerals. Received: 23 November 1998 / Accepted: 11 October 1999     

Phase transition and mixing behaviour of the cummingtonite–grunerite solid solution

 Natural amphiboles with composition close to the binary join cummingtonite–grunerite and crystals of the same samples annealed at 700 °C for 55.5 h, in order to obtain different degrees of non-convergent cation order, have been characterised by means of X-ray single-crystal diffraction and IR spectroscopy. Long-range order parameters describing the non-convergent order of Mg/Fe among the different octahedral sites have been calculated from the site occupancies of the investigated samples. Values of the O6-O5-O6 angles and of the 〈M4-O〉 mean bond distances depend on the C2/m → P2₁/m phase transition for a given degree of order. In the IR spectra, only two phonon lines dominated by the bending of the tetrahedral chains are sensitive to the displacive phase transition and to the different degree of cation order; all the other wavenumber shifts are correlated with compositional changes only. The local strains arising from the cation substitution, ordering and phase transition have been quantified by means of the autocorrelation function. Very small local heterogeneities are associated with the Mg/Fe substitution and disordering in samples at intermediate composition. The displacive phase transition seems to occur in order to reduce local distortions and the P2₁/m samples are as homogeneous as orthorhombic anthophyllites. The orthorhombic structure, however, appears less flexible than the monoclinic in accommodating cations larger than Mg at the octahedral sites. Received: 9 February 2000 / Accepted: 30 September 2000     

Simulation studies on the pyrope-grossular garnet solid solution

The local structural response of Ca/Mg substitution and the energetic effects associated with dodecahedral ordering in the pyrope-grossular garnet solid solution are derived from a combination of static lattice energy calculations and Monte Carlo simulations. We start with a thorough analysis of the goodness of the empirical potential models used for the modelling of aluminosilicate garnets. The degree of polyhedral distortion was found to be a sensitive indicator for the quality of the model and, by comparison with experimental data, was used to select the best of several available empirical potentials. The Ca/Mg substitution on the dodecahedral site in garnet was found to produce strong local distortions in the surrounding tetrahedral and octahedral polyhedra. This arises from the absence of rigid unit modes (RUMS) in the garnet structure, because local rotations of otherwise rigid SiO₄ tetrahedra and AlO₆ octahedra cannot occur in order to accommodate different-sized divalent cations in the dodecahedral sites. Strain effects, therefore, mainly govern the dodecahedral substitution, and the corresponding strain field around a dodecahedral site has a minimum radius of 5?Å. Pyrope-grossular solid solution compositions were modelled using a supercell approach. For several garnet compositions many different configurations representing individual disordered arrangements were relaxed. The resulting energies were analyzed in terms of different-neighbour interactions to determine the parameters of a model Hamiltonian. The corresponding interaction energies were found to be virtually independent of composition. Surprisingly, the nearest-neighbour interaction between edge-sharing dodecahedra is of no particular significance in the garnets. Instead, the strongest interaction is only via the third-nearest neighbours, i.e. dodecahedra that are edge-shared to a common SiO₄ tetrahedron. This cannot lead to dodecahedral long-range order in garnets, but can produce significant amounts of short-range order. Monte Carlo simulations were performed on several compositions to determine the macroscopic effects such as NMR-based cluster occupancy, ordering energy and configurational entropy of the short-range ordering process. As expected, the samples tend to random disorder at high temperatures, and at low temperatures it is compositions nearer Py50Gr50 that depart most strongly from random mixing. For example, a maximum reduction of 3.5?J?mol^?1?K^?1 is predicted for Py75Gr25 and ～10?J?mol^?1?K^?1 for Py50Gr50. A comparison of NMR cluster occupancy with experimental ²⁹Si MAS NMR resonance intensity is partly successful. However, the changes in NMR cluster occupancy are relatively low (～5%) compared to changes in configurational entropy (～30%), implying that it might be difficult to estimate exact entropy data from ²⁹Si MAS NMR line intensities.     

Structural relaxation in the MnCO3–CaCO3 solid solution: a Mn K-edge EXAFS study

Mn K-edge EXAFS spectroscopy of solid-solution samples encompassing the complete MnCO₃–CaCO₃ series shows that first-shell Mn–O distances deviate little from the 2.19-Å distance observed in pure MnCO₃. Very slight lengthening is observed only in the limiting case of dilute Mn(II) calcite solid solutions, where the Mn–O distance is 2.21 Å. The observed nearly complete structural relaxation and the composition independence of the Mn–O distance are consistent with the Pauling model behavior of solid solutions, and agree with previous studies showing a high degree of relaxation around hetero-sized substituents in the calcite structure. Strain occurs through bond bending, which is facilitated by the exclusively corner-sharing topology of calcite. Observed distances from Mn to more distant neighbors show significant variation across the solid-solution series that resembles Vegard's law-type behavior but reflects averaging. The high degree of relaxation suggests modest enthalpies of mixing in the solution, consistent with calorimetric studies.     

Lattice vibrational modes in synthetic tremolite-Sr-tremolite and tremolite-richterite solid solutions

Powder IR spectra of synthetic richterite-tremolite and Sr-tremolite-tremolite solid solutions were obtained in the spectral range between 1400 and 600?cm^?1. Under the consideration of the crystal structure and the Wykoff positions of the atoms in the primitive unit cell, the number, type and symmetry of vibrational modes were deduced. The space group of tremolite C_2h was used as the factor group leading to 16 theoretical stretching vibrations in the IR range caused by the Si₄O₁₁ ^∞-ribbon. The energy of the internal vibrations of the Si₄O₁₁ ^∞-ribbon is a function of the relative bond strengths and masses of nearby ions. For the amphiboles a one-mode behavior was observed for all the Si-O, Si-O-Si and O-Si-O stretching vibrations, indicating no clustering in the two solid solution series. In both solid solution series the vibrational energy of the stretching vibrations is a linear function of composition. In the system richterite-tremolite a shift of the stretching frequencies of the Si₄O₁₁ ^∞-ribbon over the whole compositional range of up to 30?cm^?1 was observed. In contrast, for Sr-tremolite-tremolite the maximum shift was only 5?cm^?1. These quite small band shifts allow the (Si₄O₁₁)^∞-ribbon to be treated as an isolated entity for factor group analysis. Nevertheless, by the two exchange mechanisms, Ca(M4)???Sr(M4) and □(A) Ca(M4)???Na(A)Na(M4), the FWHHs increased and the amplitudes decreased, indicating a slight distortion of the ribbon. For Sr-tremolite-tremolite only a linear expansion of the lattice was observed. In the series richterite-tremolite individual bond angles of the SiO₄ tetrahedra are additionally changed, causing the higher energy shift of the bands. The strongest and sharpest bands were observed for the end member tremolite. The one-mode behavior of the Si₄O₁₁-double chain indicates that there is no short-range order of Na/Ca and Ca/Sr at the M4 sites of these amphiboles.     

Electric conductivity of Fe2SiO4–Fe3O4 spinel solid solutions

 The spinel solid solution was found to exist in the whole range between Fe₃O₄ and γ-Fe₂SiO₄ at over 10 GPa. The resistivity of Fe_{3− x} Si _x O₄ (0.0<x<0.288) was measured in the temperature range of 80∼300 K by the AC impedance method. Electron hopping between Fe³⁺ and Fe²⁺ in the octahedral site of iron-rich phases gives a large electric conductivity at room temperature. The activation energy of the electron hopping becomes larger with increasing γ-Fe₂SiO₄ component. A nonlinear change in electric conductivity is not simply caused by the statistical probability of Fe³⁺–Fe²⁺ electron hopping with increasing the total Si content. This is probably because a large number of Si⁴⁺ ions occupies the octahedral site and the adjacent Fe²⁺ keeping the local electric neutrality around Si⁴⁺ makes a cluster, which generates a local deformation by Si substitution. The temperature dependence of the conductivity of solid solutions indicates the Verwey transition temperature, which decreases from 124(±2) K at x=0 (Fe₃O₄) to 102(±5) K at x=0.288, and the electric conductivity gap at the transition temperature decreases with Si⁴⁺ substitution. Received: 15 March 2000 / Accepted: 4 September 2000     

Magnetic properties of the Fe2SiO4-Fe3O4 spinel solid solutions

 Magnetic measurement of Fe_{3− x} Si _x O₄ spinel solid solutions indicates that their Curie temperatures decrease gradually, but not linearly, from 851 to 12 K with increasing content of nonmagnetic ions Si⁴⁺. Magnetic hysteresis becomes more noticeable in solid solutions having a larger content of Fe₂SiO₄. Saturation magnetizations of Fe_{3− x} Si _x O₄ samples increase up to x=0.357 and they are easily saturated in the field of H=0.1 T. However, magnetization of the sample of x=0.794 does not approach saturation even at high field of H=7.0 T and has a large coercive force. The Si⁴⁺ disordered distribution is confirmed to be ^tetr[Fe³⁺ _{1− x + x t} Si⁴⁺ _{x (1− t )}] ^octa[Fe²⁺ _{1+ x} Fe³⁺ _{1− x − x t} Si⁴⁺ _{x t} ] O₄ by the spin moment, which is consistent with site occupancy obtained from X-ray crystal structure refinement. Their molecular magnetizations would be expressed as M _B={4(1+x)+10xt}μ_B as functions of composition parameter x and Si⁴⁺ ordering parameter t of the solid solution. The sample of x=0.794 is antiferromagnetic below the Néel temperature, mainly due to the octahedral cation interaction M _O–M _O, while both M _T–M _O and M _O–M _O interactions induce a ferrimagnetic property. Concerning magnetic spin configuration, in the case of x>0.42, the lowest dɛ level becomes a singlet, resulting in no orbital angular momentum. Received: 20 April 2000 / Accepted: 11 September 2000     

Cation distribution and structure modelling of spinel solid solutions

 This paper presents an improved generalisation of cation distribution determination based on an accurate fit of all crystal-chemical parameters. Cations are assigned to the tetrahedral and octahedral sites of the structure according to their scattering power and a set of bond distances optimised for spinel structure. A database of 295 spinels was prepared from the literature and unpublished data. Selected compositions include the following cations: Mg²⁺, Al³⁺, Si⁴⁺, Ti⁴⁺, V³⁺, Cr³⁺, Mn²⁺, Mn³⁺, Fe²⁺, Fe³⁺, Co²⁺, Ni²⁺, Zn²⁺ and vacancies. Bond distance optimisation reveals a definite lengthening in tetrahedral distance when large amounts of Fe³⁺ or Ni²⁺ are present in the octahedral site. This means that these cations modify the octahedral angle and hence the shared octahedral edge, causing an increase in the tetrahedral distance with respect to the size of the cations entering it. Some applications to published data are discussed, showing the capacity and limitations of the method for calculating cation distribution, and for identifying inconsistencies and inaccuracies in experimental data. Received: 19 February 2001 / Accepted: 1 June 2001     

Atmospheric oxidation of the pyritic waste rock in Maardu, Estonia, 2: an assessment of aluminosilicate buffering potential

 The assessment of the aluminosilicate buffering potential during acid weathering of the Estonian alum shale is provided. It is found that the stoichiometric interaction between dissolved pyrite oxidation products and illite of the shale best describe the buffering process and are consistent with earlier field studies. The scheme includes incongruent dissolution of illite with smectite and K-jarosite precipitating. This complex mechanism involves buffering of 8% of the acidity by K⁺ and temporary precipitation of 25% of the acidity as K-jarosite. Dissolution proceeds at a low pH (1.5–3) until all pyrite in the shale particle is oxidised. Hence, if the total amount of illite present is larger than needed for stoichiometric interactions, only part of it is involved in a buffering process, neutralising a certain percentage of acidity. The next stage in shale weathering is the incongruent dissolution of K-jarosite with the release of the precipitated acidity and the formation of ferric oxyhydroxide. Received: 3 August 1998 · Revised paper: 26 January 1999 · Accepted: 23 February 1999     

Low-temperature thermodynamic properties of disordered zeolites of the natrolite group

 The heat capacity of paranatrolite and tetranatrolite with a disordered distribution of Al and Si atoms has been measured in the temperature range of 6–309 K using the adiabatic calorimetry technique. The composition of the samples is represented with the formula (Na_1.90K_0.22Ca_0.06)[Al_2.24Si_2.76O₁₀]·nH₂O, where n=3.10 for paranatrolite and n=2.31 for tetranatrolite. For both zeolites, thermodynamic functions (vibrational entropy, enthalpy, and free energy function) have been calculated. At T=298.15 K, the values of the heat capacity and entropy are 425.1 ± 0.8 and 419.1 ±0.8 J K⁻¹ mol⁻¹ for paranatrolite and 381.0 ± 0.7 and 383.2 ± 0.7 J K⁻¹ mol⁻¹ for tetranatrolite. Thermodynamic functions for tetranatrolite and paranatrolite with compositions corrected for the amount of extraframework cations and water molecules have also been calculated. The calculation for tetranatrolite with two water molecules and two extraframework cations per formula yields: C _p (298.15)=359.1 J K⁻¹ mol⁻¹, S(298.15) −S(0)=362.8 J K⁻¹ mol⁻¹. Comparing these values with the literature data for the (Al,Si)-ordered natrolite, we can conclude that the order in tetrahedral atoms does not affect the heat capacity. The analysis of derivatives dC/dT for natrolite, paranatrolite, and tetranatrolite has indicated that the water- cations subsystem within the highly hydrated zeolite may become unstable at temperatures above 200 K. Received: 30 July 2001 / Accepted: 15 November 2001     

Computer simulation of defect structure in sillimanite and mullites

 Defect structure and the defect formation in mullites and sillimanites have been investigated using computer simulation techniques. From point defect chemistry and computer simulations, oxygen vacancies are identified as the majority defect responsible for oxygen transport in mullite. The defect formation energies are between 3 and 4 eV. Using supercell calculation methods, the random structure is identified to be stable in mullites, whereas the ordered structure can be confirmed to be more stable in sillimanite. An energy of 0.7–1.0 eV for the association of oxygen vacancies with cations is estimated. Received: 11 May 2001 / Accepted: 12 December 2001     

Integrated use of hydrochemistry and resistivity methods in groundwater contamination caused by a recently closed solid waste site

 In order to investigate the groundwater contamination by solid waste disposal using both hydrochemical and geophysical methods, the Halkalı (I˙stanbul) solid waste disposal site which was closed in 1994 was investigated. The disposal site lies on a ridge between two valleys filled with alluvium. A total of six boreholes were drilled on two lines across the Menekşe valley adjacent to the Halkalı site. Groundwater samples collected from these boreholes were analyzed for various contaminant parameters. The results indicate that TDS and chloride concentrations decrease horizontally away from the waste site whereas they increase with depth. Electrical soundings carried out at 12 locations yielded high resistivity values at the upstream part of Menekşe valley while lower values were obtained from the locations near the leachate seepage points. Received: 11 November 1997 · Accepted: 23 February 1998     

High-temperature heat capacity and thermodynamic properties for end-member titanite (CaTiSiO5)

 The heat capacity of end-member titanite and (CaTiSiO₅) glass has been measured in the range 328–938 K using differential scanning calorimetry. The data show a weak λ-shaped anomaly at 483 ± 5 K, presumably associated with the well-known low-pressure P2₁/a ⇆ A2/a transition, in good agreement with previous studies. A value of 0.196 ± 0.007 kJ mol⁻¹ for the enthalpy of the P2₁/a ⇆ A2/a transition was determined by integration of the area under the curve for a temperature interval of 438–528 K, bracketing the anomaly. The heat capacity data for end-member titanite and (CaTiSiO₅) glass can be reproduced within <1% using the derived empirical equations (temperature in K, pressure in bars):

Thermodynamics and crystal chemistry of the hematite–corundum solid solution and the FeAlO3 phase

High-temperature oxide-melt calorimetry and Rietveld refinement of powder X-ray diffraction patterns were used to investigate the energetics and structure of the hematite–corundum solid solution and ternary phase FeAlO₃ (with FeGaO₃ structure). The mixing enthalpies in the solid solution can be described by a polynomial ΔH_mix=WX _hem(1?X _hem) with W=116 ± 10 kJ mol^?1. The excess mixing enthalpies are too positive to reproduce the experimental phase diagram, and excess entropies in the solid solution should be considered. The hematite–corundum solvus can be approximately reproduced by a symmetric, regular-like solution model with ΔG _excess=(W _H ?TW _S )X _hem X _cor, where W _H= 116 ± 10 kJ mol^?1 and W _S =32 ± 4 J mol^?1 K^?1. In this model, short-range order (SRO) of Fe/Al is neglected because SRO probably becomes important only at intermediate compositions close to Fe:Al=1:1 but these compositions cannot be synthesized. The volume of mixing is positive for Al-hematite but almost ideal for Fe-corundum. Moreover, the degree of deviation from Vegard's law for Al-hematite depends on the history of the samples. Introduction of Al into the hematite structure causes varying distortion of the hexagonal network of oxygen ions while the position of the metal ions remains intact. Distortion of the hexagonal network of oxygen ions attains a minimum at the composition (Fe_0.95Al_0.05)₂O₃. The enthalpy of formation of FeAlO₃ from oxides at 298 K is 27.9 ± 1.8 kJ mol^?1. Its estimated standard entropy (including configurational entropy due to disorder of Fe/Al) is 98.9 J mol^?1 K^?1, giving the standard free energy of formation at 298 K from oxides and elements as +19.1 ± 1.8 and ?1144.2 ± 2.0 kJ mol^?1, respectively. The heat capacity of FeAlO₃ is approximated as C _p (T in K)= 175.8 ? 0.002472T ? (1.958 × 10⁶)/T ²? 917.3/T ^0.5+(7.546 × 10^?6) T ² between 298 and 1550 K, based on differential scanning calorimetric measurements. No ferrous iron was detected in FeAlO₃ by Mössbauer spectroscopy. The ternary phase is entropy stabilized and is predicted to be stable above about 1730 ± 70 K, in good agreement with the experiment. Static lattice calculations show that the LiNbO₃-, FeGaO₃-, FeTiO₃-, and disordered corundum-like FeAlO₃ structures are less stable (in the order in which they are listed) than a mechanical mixture of corundum and hematite. At high temperatures, the FeGaO₃-like structure is favored by its entropy, and its stability field appears on the phase diagram.