Munakataite, Pb2Cu2(Se4+O3)(SO4)(OH)4, has been found in association with a variety of rare secondary Te minerals at Otto Mountain, San Bernardino County, California, USA. It is very rare and occurs as subparallel bundles of blue needles up to 1 mm long. Electron microprobe analyses provided the empirical formula Pb1.96Cu1.60[(Se0.894+S0.11)∑1O3](SO4)[(OH)3.34(H2O)0.66]∑4- Munakataite is mono-clinic, space group P21/m, with cell parameters a = 9.8023(26), b = 5.6751(14), c = 9.2811(25) Å, β = 102.443(6), V = 504.2(2) Å3 and Z = 2. The crystal structure, solved by direct methods and refined to R1 = 0.0308 for 544 Fo > 4σF reflections, consists of Jahn-Teller-distorted Cu2+O6 square bipyramids. which form chains along b by sharing trans edges across their square planes. The chains are decorated by SO4 tetrahedra and Se4+O3 pyramids, which bond to apical corners of adjacent bipyramids. The chains are linked to one another via bonds to two different PbO9 polyhedra, only one of which exhibits one-sided coordination typical of Pb2+ with a stereochemically active 6s2 lone-electron-pair. Munakataite is isostructural with schmiederite and the structure is closely related to that of linarite.

Bendadaite, ideally Fe2+Fe23+ (AsO4)2(OH)2·4H2O, is a new member of the arthurite group. It was found as a weathering product of arsenopyrite on a single hand specimen from the phosphate pegmatite Bendada, central Portugal (type locality). Co-type locality is the granite pegmatite of Lavra do Almerindo (Almerindo mine), Linópolis, Divino das Laranjeiras county, Minas Gerais, Brazil. Further localities are the Veta Negra mine, Copiapó province, Chile; Oumlil-East, Bou Azzer district, Morocco; and Pira Inferida yard, Fenugu Sibiri mine, Gonnosfanadiga, Medio Campidano Province, Sardinia, Italy.

Type bendadaite occurs as blackish green to dark brownish tufts (<0.1 mm long) and flattened radiating aggregates, in intimate association with an intermediate member of the scorodite–mansfieldite series. It is monoclinic, space group P21/c, with a = 10.239(3) Å, b = 9.713(2) Å, c = 5.552(2) Å, β = 94.11(2)°, V = 550.7(2) Å3, Z = 2. Electron-microprobe analysis yielded (wt.%): CaO 0.04, MnO 0.03, CuO 0.06, ZnO 0.04, Fe2O3 (total) 43.92, Al2O3 1.15, SnO2 0.10, As2O5 43.27, P2O5 1.86, SO3 0.03. The empirical formula is (Fe2+0.52Fe3+0.32☐0.16)Σ1.00(Fe3+1.89Al0.11)Σ2.00(As1.87P0.13)Σ2.00O8(OH)2.00·4H2O based on 2(As,P) and assuming ideal 8O, 2(OH), 4H2O and complete occupancy of the ferric iron site by Fe3+ and Al. Optically, bendadaite is biaxial, positive, 2Vest. = 85±4°, 2Vcalc. = 88°, with α 1.734(3), β 1.759(3), γ 1.787(4). Pleochroism is medium strong: X pale reddish brown, Y yellowish brown, Z dark yellowish brown; absorption Z > Y > X, optical dispersion weak, r > v. Optical axis plane is parallel to (010), with X approximately parallel to a and Z nearly parallel to c. Bendadaite has vitreous to sub-adamantine luster, is translucent and non-fluorescent. It is brittle, shows irregular fracture and a good cleavage parallel to {010}. Dmeas. 3.15±0.10 g/cm3, Dcalc. 3.193 g/cm3 (for the empirical formula). The five strongest powder diffraction lines [d in Å (I)(hkl)] are 10.22 (10)(100), 7.036 (8)(110), 4.250 (5)(111), 2.865 (4)(), 4.833 (3)(020,011). The d spacings are very similar to those of its Zn analogue, ojuelaite. The crystal structure of bendadaite was solved and refined using a crystal from the co-type locality with the composition (Fe2+0.95☐0.05)Σ1.00(Fe3+1.80Al0.20)Σ2.00(As1.48P0.52)Σ2.00O8(OH)2·4H2O (R = 1.6%), and confirms an arthurite-type atomic arrangement.

The crystal structure of byzantievite, Ba5(Ca,REE,Y)22(Ti,Nb)18(SiO4)4[(PO4),(SiO4)]4 (BO3)9O21[(OH),F]43(H2O)1.5, a new mineral from the moraine of the Dara-i-Pioz glacier, the Alai mountain ridge, Tien-Shan Mountains, northern Tajikistan, was solved by direct methods and refined to R1 = 13.14% based on 3794 observed [Fo >4σ|F|] unique reflections measured with Mo-Kα X-radiation on a Bruker P4 diffractometer equipped with a CCD detector. Byzantievite is hexagonal, space group R3, a = 9.1202(2) Å, c = 102.145(5) Å, V = 7358.0(5) Å3, Z = 3, Dcalc. = 4.151 g cm–3. The empirical formula (electron microprobe analysis) is Ba5.05[(Ca8.99Sr0.96Fe2+0.42Na0.20)Σ10.57(Ce3.46La1.54Nd1.20Pr0.30Sm0.26Dy0.41Gd0.32Th0.39U4+0.17)Σ8.05Y3.53]Σ22.15(Ti12.31Nb5.30)Σ17.61(SiO4)4.65(PO4)3.12(BO3)8.89O22.16(OH)38.21F4.89(H2O)1.5, Z = 3, calculated on the basis of 124.5 (O + F) a.p.f.u. The H2O and OH contents were calculated from structure refinement (F + OH = 43 a.p.f.u.; H2O = 1.5 a.p.f.u..), and B was determined by SIMS. The crystal structure is a framework of Ti-Ba-Ca-REE-dominant polyhedra and SiO4, PO4 and BO3 groups. In the crystal structure, there are 50 cation sites, 23 of which are fully occupied and 27 partly occupied: six of the 27 partly-occupied sites are >50% occupied, 21 <50% occupied. The crystal structure of byzantievite is an intercalation of three components, one fully ordered with 100% occupancy of cation sites, and two partly ordered with cation-site occupancies of 67% and 17% respectively. Byzantievite is the only known mineral that contains all three of the oxyanions (BO3), (SiO4) and (PO4) as essential components.

The crystal structure of H3O+-exchanged pharmacosiderite (pharmacosiderite is KFe4(AsO4)3(OH)4·nH2O, sensu stricto) has been determined by single-crystal X-ray diffraction and refined to R1 = 0.0418. H3O+-exchanged pharmacosiderite, (H3O+)Fe4(AsO4)3(OH)4·4.5H2O, is cubic, space group Pm, with a = 7.982(9) Å, V = 508.5(9) Å3 and Z = 1. The structure broadly conforms to that of the general pharmacosiderite structure type, with the hydronium ion generated by partial protonation of a site corresponding to a molecule of water of crystallization and its symmetry-related equivalents. In addition, the structure of a “pharmacosiderite” from Cornwall, United Kingdom, in which no alkali metals could be detected, has been re-evaluated and found to be consistent with that of the H3O+- exchanged structure. Its composition is (H3O+)Fe4(AsO4)3(OH)4·4H2O, with the partially occupied water found for the exchanged structure at (½, ½, ½) being absent in this case.

New data on rare earth element (REE) concentrations and distribution patterns of columbite-tantalite minerals from Ta-ore provinces worldwide are presented. The REE geochemistry was studied by laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS). Five major types of chondrite-normalized REE distribution patterns are defined for columbite-group minerals (CGM) from lithium-caesium-tantalum (LCT) pegmatites and rare-metal granites. Features to discriminate between the types include (1) the shape of the pattern (e.g. flat or concave), (2) calculated ratios between groups of the REE (e.g. heavy REEN/middle REEN), and (3) the presence and intensity of anomalies (e.g. Ce*, Eu*). Four pegmatites in central and southern Africa are used as case studies to discuss application of the types of REE patterns in individual deposits. The REE fractionation during progressive evolution of the melt in a pegmatite body (either Nb → Ta or Fe → Mn fractionation lines, or both) results in smaller heavy REEN/middle REEN ratios whereas replacement of primary CGM by secondary CGM produces modifications in the light REEN patterns and the heavy REEN/middle REEN ratios also. Critical features of REE patterns such as highly variable heavy REEN/middle REEN ratios or striking differences in the appearance of Eu anomalies are discussed considering structural data of the host minerals and the differentiation behaviour of the pegmatitic melt. In general, CGM from each individual Ta-ore province are characterized by a predominance of one type of REE distribution pattern. Consequently, these patterns are suitable for tracing the origin of tantalum ore concentrates (e.g. as a forensic tool).

Atoll garnet has been found in metabasites and quartz- and mica-rich rocks that have experienced low- to medium-temperature, high-pressure eclogite facies metamorphism in the Krušné Hory (Erzgebirge). They occur in several localities but are restricted to thin, texturally distinct zones, even on a thin-section scale. The mechanism of atoll garnet formation is documented by a series of micrographs and compositional maps and profiles of atoll garnet in combination with textural relations to other phases in the rocks. The core of full garnet or its relics in the atoll garnet have larger Ca and Fe, but smaller Mg contents, compared with the thin rim (ring). In addition to quartz, Na-Ca amphibole and phengite, the atoll cores are filled by a new garnet that has a composition similar to the outer rim. Formation of the atoll garnet is interpreted as resulting from fluid infiltration and element exchange between the garnet core and matrix, a process facilitated by a temperature increase during eclogite facies metamorphism. In addition to fluid access, the primary textures, mainly grain size, were also effective for the atoll garnet formation. Small grain fractions with thin rims were easily infiltrated by fluid, which used the short distance for element exchange between core and matrix. The core garnet was gradually dissolved and replaced by new garnet having the same crystallographic orientation as the rim or relics in the core.

The structure of hibonite from Myanmar (space group P63/mmc, Z = 2, at room temperature a = 5.5909(1), c = 21.9893(4) Å), with simplified formula CaAl12O19 and composition (Ca0.99Na0.01)Σ1.00 was investigated between temperatures of 100 K and 923 K by single-crystal X-ray diffraction methods. Structure refinements have been performed at 100, 296, 473 and 923 K. In hibonite from Myanmar, Ti substitutes for Al mainly at the octahedral Al4 site and, to a lesser degree, at the trigonal bipyramidal site, Al2. The Al4 octahedra build face-sharing dimers. If Ti4+ substitutes at Al4, adjacent cations repulse each other for electrostatic reasons, leading to off-centre cation displacement associated with significant bond-length distortion compared to synthetic (Ti-free) CaAl12O19. Most Mg and smaller proportions of Zn and Si are assigned to the tetrahedral Al3 site. 12-coordinated Ca in hibonite replaces oxygen in a closest-packed layer. However, Ca is actually too small for this site and engages in a ‘rattling-type’ motion with increasing temperature. For this reason, Ca does not significantly increase thermal expansion coefficients of hibonite. The expansion of natural Ti,Mg-rich hibonite between 296 and 923 K along the x and the z axes is αa = 7.64×10–6 K–1 and αc = 11.19×10–6 K–1, respectively, and is thus very similar to isotypic, synthetic CaAl12O19 and LaMgAl11O19 (LMA).

The crystal structure of the rare mineral dalnegroite, Tl5−xPb2x(As,Sb)21−xS34 with x ≈ 1, was determined for a crystal from Lengenbach, Binn Valley, Switzerland. The structure is triclinic, space group P1, with a = 16.218(3), b = 42.546(7), c = 8.558(1) Å , α = 95.70(4), β = 90.18(3), γ = 96.38(4)°, V = 5838.9(9) Å3, Z = 4. Refinement of an isotropic model led to an R1 index of 0.0536 for 22226 observed reflections and 980 parameters, and R1 = 0.0590 for all 25266 independent reflections. Although dalnegroite cannot be considered a layered compound, its structure can be usefully described as a regular alternation of two kinds of layers stacked along the b axis, with four layers in the unit cell: (1) a layer 7.8 Å thick, at y ≈ 0.15 and 0.65, can be considered as derived from the SnS archetype; (2) a layer 13.6 Å thick, at y ≈ 0.35 and 0.85, derived from the PbS archetype. Different chemical compositions, such as Tl:Pb and Sb:As ratios, for different samples belonging to the chabournéite-dalnegroite family could play a central role in controlling different degrees of order, leading to different superstructures.

Electron backscatter diffraction (EBSD) analysis of monazite requires a comparison of empirically collected electron backscatter patterns (EBSPs) with theoretical diffraction data, or ‘match units’, derived from known crystallographic parameters. Published crystallographic data derived from compositionally varying natural and synthetic monazite are used to calculate ten different match units for monazite. These match units are used to systematically index EBSPs obtained from four natural monazite samples with different compositions. Analyses of EBSD data, derived from the indexing of five and six diffraction bands using each of the ten match units for 10,000 EBSPs from each of the four samples, indicate a large variation in the ability of the different match units to correctly index the different natural samples. However, the use of match units derived from either synthetic Gd or Eu monazite crystallographic data yield good results for three of the four analysed monazites. Comparison of sample composition with published monazite compositions indicates that these match units are likely to yield good results for the EBSD analysis of metamorphic monazite. The results provide a clear strategy for optimizing the acquisition and analysis of EBSD data from monazite but also indicate the need for the collection of new crystallographic structure data and the subsequent generation of more appropriate match units for natural monazite.

The crystal chemistry of tapiolite is more complicated than previously recognized. Perusal and synthesis of relatively recent publications reveals a ‘hidden’ story about tapiolite, one that: (1) permits quantification of the effects of Mn, Fe3+, Ti (and Sn) upon the unit-cell parameters of maximally ordered tapiolite; (2) shows that the effect of Nb = Ta substitution upon unit-cell parameters is insignificant; (3) casts doubt on the usefulness of heating experiments for accurately evaluating cation order in tapiolite; (4) suggests that rutile exsolves from tapiolite upon heating in air at lower temperatures than previously established; (5) shows that the maximum solubility of TiO2 in FeTa2O6 tapiolite during the later stages of crystallization of granitic pegmatites is ∼1 mol.%; (6) shows that the heating paths for at least some natural tapiolite samples are not necessarily linear, consisting of two sequential steps; (7) sheds doubt on the usefulness of unit-cell parameters as quantitative measures of long-range cation order, Q, in tapiolite; but (8) does show that the intensity ratio I011/I110 can provide an accurate and reasonably precise measure of cation order in tapiolite.

The structure of the high-temperature form of CaSeO4, formed by dehydration of the gyspum-type structure dihydrate is presented. The material is equivalent to that described previously as a P212121 form, but is however, a monazite with unit cell a = 6.85661(16) Å, b = 7.04962(15) Å, c = 6.68817(15) Å and β = 104.2675(21)° in space group P121/n1. Also presented is evidence for two intermediate trigonal and pseudo-trigonal phases related to the structurally similar minerals rhabdophane, bassanite, and γ-CaSO4. This result permits closer comparisons between intermediate-temperature structures of the selenate with related sulphates, and orthophosphates with a view to extending structural stability via synthesis of solid-solutions.

The study presents analysis from members of the tennantite–tetrahedrite solid-solution series (‘fahlore’) from 78 locations in the Schwarzwald ore district of SW Germany. Electron microprobe analysis is used to correlate the compositional variations of the fahlores with mineral association, host rock, tectonic history and precipitation mechanisms. Results indicate that most fahlores from gneiss-hosted veins do not have distinctive geochemical characteristics and range from tetrahedrite to tennantite end-member composition with variable trace-element content. However, diagenetically formed fahlore has a near-end-member tennantite composition with very small trace-element content. Red-bed-hosted fahlore formed by fluid mixing is tennantite enriched in Hg that probably has its source in the red-bed sediments. Fahlore formed from granite-related late-magmatic fluids, or from mixing of fluids of which one has equilibrated with granitic basement rocks, is typically As- and Bi-rich (up to 22.2 wt.% Bi). Gneiss-hosted fahlore formed by fluid cooling is Ag-rich near-end-member tetrahedrite. Some fahlores reflect their paragenetic association, e.g. a large Ag content in association with Ag-bearing minerals or a large Co and Ni in association with Co- and Ni-arsenides.

Although they have similar compositions, gneiss-hosted fahlores show systematic variations in Ag contents and Fe/Zn ratios between the Central and the Southern Schwarzwald with Fe-rich fahlore in higher stratigraphic levels (North) and Zn- and Ag-rich fahlore in lower stratigraphic levels (South). We show that fahlore composition varies with precipitation mechanism (cooling vs. mixing vs. diagenesis), depth of formation, paragenetic association and host rock. Comparison with fahlores from other European occurrences indicates that these conclusions are consistent with fahlore systematics found elsewhere, and could be used to infer details of ore-forming processes.

Kirschsteinite (mainly CaFeSiO4), in association with forsterite-fayalite solid solutions, melilite, clinopyroxene, leucite, nepheline and several minor phases, is reported for the first time in the 290 ka Capo di Bove melilite-leucitite lava flow of the Alban Hills volcanic complex (Italy). Kirschsteinite appears very late in the crystallization sequence, as indicated by textural and chemical evidence (Mg# ≈ 25). It shows significant solid solution towards fayalite, and is found with groundmass phases indicating a relatively low-temperature melt composition more silica-undersaturated than phonolite (i.e. devoid of alkali feldspar). This clearly points to the very silica-undersaturated nature of almost all the Alban Hills volcanic rocks.

Data on the dehydration of pentahydroborite, CaB2O(OH)6·2H2O and nifontovite, Ca3B6O6(OH)12·2H2O from the Fuka mine, Japan are presented. Critical temperatures of the dehydration of the borates were determined by thermogravimetric analysis/differential thermal analysis measurements. The untreated mineral samples and their heating products were investigated by X-ray diffraction and Raman spectroscopy. Upon dehydration, both minerals decompose and undergo amorphization, and at greater temperatures crystallize as an orthorhombic calcium borate, CaB2O4 (Pnca). The dehydration paths of the two minerals are different, with nifontovite showing a greater resistance to decomposition and amorphization than pentahydroborite. Differences in the dehydration processes are related to the residuals of the water content and structural accommodation of the borate polyanion.

Ammonioalunite, ammoniojarosite and their solid-solution series found on a burning coal dump at Czerwionka, Upper Silesian Coal Basin, Poland, were examined using powder X-ray diffraction, electron probe microanalysis, inductively coupled plasma mass spectrometry, infrared spectroscopy and thermal analysis methods. The minerals occur as yellow, fine-grained crusts and botryoidal masses in the outer part of a sulphate crust found ∼1 m below the surface. The crust is composed mainly of godovikovite and tschermigite that formed by interaction of sour fire gases or solutions and waste materials beneath the surface of the burning coal dump at temperatures of at least 80—100°C. The crystals often reveal oscillatory zoning due to different Al and Fe contents in thin bands, from near end-members to extensive solid solutions. Our analyses suggest the existence in nature of a complete solid solution between ammonioalunite and ammoniojarosite. They also carry essential amounts of chlorine.

The collapse of minerals and mineral assemblies under external stress is modelled using a master curve where the stress failure is related to the relative, effective elastic moduli which are in turn related to the porosity of the sample. While a universal description is known not to be possible, we argue that for most porous materials such as shales, silica, cement phases, hydroxyapatite, zircon and also carbonates in corals and agglomerates we can estimate the critical porosity ϕc at which small stresses will lead to the collapse of the sample. For several samples we find ϕc ~0.5 with an almost linear decay of the bulk moduli with porosity at ϕc <0.5. The second scenario involves the persistence of elasticity for porosities until almost 1 whereby the bulk modulus decreases following a power law κ ~ (1–ϕm, m>2, between ϕ = 0.5 and ϕ = 1.

On 27 January 2010 a small eruption of basaltic lava occurred 75 km NE of Quetta, Pakistan. This was highly unusual – no eruptions from this magmatically inactive area have ever been reported. Two petrographically distinct basalts types were indentified in the vesicular eruptive products. One basalt type consists of completely fresh, light brown glass with a few (<1 vol.%) partially resorbed quartz-rich xenoliths, whereas the other type is non-glassy and the lava is completely devitrified. These types also have slightly different geochemical signatures that can be partially explained by crustal assimilation. Re-melting of local basaltic rocks by short circuiting of a ruptured high-tension electrical cable is considered unlikely. Mantle melt modelling suggests that the lavas have been largely derived from a source in the garnet-spinel transition zone, i.e. well within the lithosphere. It is proposed that localized asthenospheric melting resulted in relatively depleted melts which were substantially contaminated by fusible lithospheric mantle en route to the surface. Further small-scale eruptions cannot be ruled out.

The temperature (T) and pressure (P) dependence of dielectric and conductivity properties of natural leucite were determined using complex impedance spectroscopy at frequencies from 103 to 106 Hz. Experiments were carried out in a Walker multi-anvil cell at 1 atm and P from 2.5 to 6 GPa and at T from 350 to 800°C. At pressure >6 GPa and temperature >790°C the leucite broke down to kalsilite+sanidine and dielectric properties for this phase assemblage are given at 6.0–7.0 GPa and T to 1050°C.

Leucite conductivity increases with increasing T and decreases with increasing P reflecting their different effects on migration of K cations within the channels in the leucite aluminosilicate framework. Activation energies for K+ migration in leucite increase with increasing pressure (0.74–0.97 eV; 70.0–93.2 kJ/mol) and activation volumes for leucite increase with increasing T (6.42–9.51 cm3/mol; 400–700°C). The latter data provide model K+ cation diameters increasing from 2.7 Å at 400°C to 3.2 Å at 700°C. These values are consistent with the earlier suggestion of Palmer and Salje that the ionic mobility mechanism consists of diffusion along <110> rather than along the main channels parallel to <111>.

Demicheleite-(I), ideally BiSI, is the iodine-dominant analogue of demicheleite-(Br) and demicheleite-(Cl). It was found in an active medium-temperature intracrateric fumarole at La Fossa crater, Vulcano Island, Aeolian archipelago, Sicily, Italy. The mineral is the first bismuth sulphoiodide so far discovered in a wholly natural environment, and corresponds to the already known synthetic compound. It occurs as acicular to stout, translucent crystals up to 0.25 mm long in an altered pyroclastic breccia, together with demicheleite-(Br), bismoclite, bismuthinite, godovikovite, panichiite, aiolosite, brontesite, adranosite and other new phases under study. The colour is dark red to black, the lustre submetallic. The unit cell is orthorhombic, space group Pnam, with a = 8.4501(7) Å, b = 10.1470(9) Å , c = 4.1389(4) Å , V = 354.88(4) Å3, and Z = 4. The crystal habit is prismatic, with the main forms {110} and {111} inferred from analogy with demicheleite-(Br). Twinning was not observed. The strongest 6 lines in the X-ray powder diffraction pattern [dobs.(Å) (I/I0) (hkl)] are: 6.490 (100) (110); 4.346 (94) (120); 3.896 (90) (210); 2.709 (60) (310); 2.161 (38) (330); 3.243 (22) (220). The chemical analysis obtained by WDS electron microprobe gave: Bi 58.32, S 9.43, I 23.69, Br 5.66, Cl 1.01, totalling 98.11 wt.%, corresponding to an empirical formula (based on 3 a.p.f.u.) of: Bi0.97S1.03(I0.65Br0.25Cl0.10)Σ1.00. The unit-cell data are close to those of the synthetic compound, whose crystal structure is already known. The calculated density is 6.411 g cm–3.

The sodium-magnesium hydrated double salt konyaite, Na2Mg(SO4)2·5H2O, has been studied by single-crystal X-ray diffraction and thermogravimetry on synthetic samples and by quantitative X-ray diffraction utilizing the Rietveld method on natural samples from the Mount Keith mine, Western Australia. Konyaite crystallizes in space group P21/c, with the cell parameters: a = 5.7594(10), b = 23.914(4), c = 8.0250(13) Å, β = 95.288(9)°, V = 1100.6(3) Å3 and Z = 4. The crystal structure has been refined to R1 = 3.41% for 2155 reflections [Fo>4σ(Fo)] and 6.44% for all 3061 reflections, with all atoms located.

Quantitative phase analysis utilizing the Rietveld method was undertaken on five samples of konyaite-bearing mine tailings from the Mount Keith Nickel Mine, Western Australia. Konyaite was found to decompose over time and after 22 months had transformed to other sulphate and amorphous phases. Blödite did not increase in any ofthe samples indicating that konyaite may not always transform to blödite. Over the same time frame, synthetic konyaite completely decomposed to a mixture of thenardite (Na2SO4), hexahydrite (MgSO4·6H2O), blödite (Na2Mg(SO4)2·4H2O) and löweite (Na12Mg7(SO4)13). Detection of konyaite and other Mg-rich sulphates is important in terms of CO2 fixation. Magnesium bound to sulphate mineral phases reduces the overall potential of tailings piles to lock up atmospheric carbon in Mg carbonates, such as hydromagnesite. Amorphous sulphates are also highly reactive and may contribute to acid mine drainage ifpresent in large quantities, and may dissolve carbonate phases which have already sequestered carbon.