Cation Substitution in Uranyl Phosphates of the Autunite Group: Equilibrium Relations and Crystallization Between Metatorbernite and Metauranocircite

Home , Autunite, Metatorbernite

Versão online: http://www.lneg.pt/iedt/unidades/16/paginas/26/30/208 Comunicações Geológicas (2015) 102, Especial I, 27-30 ISSN: 0873-948X; e-ISSN: 1647-581X

Cation substitution in uranyl phosphates of the autunite group: equilibrium relations and crystallization between metatorbernite and metauranocircite

Substituição catiónica em fosfatos de uranilo do grupo da autunite: relações de equilíbrio e cristalização entre metatorbernite e metauranocircite

M. Andrade1, J. Duarte1, I. Martins 1, J. Reis 1, J. Mirão3, M. A. Gonçalves1,2*

Abstract: Uranyl phosphate minerals play an important role in the 1. Introduction uranium immobilization within weathering and supergene enrichment profiles. This work consists on the morphological, structural and Uranyl phosphate minerals are major constituents in weathered U chemical characterization of natural and synthetic minerals of Cu and Ba deposits and can display a multi-stage evolving history in the – metatorbernite and metauranocircite, respectively. SEM imaging has environment they crystalize. Their importance is two-fold: as revealed an extended range of morphologies, from tabular to rosette-like main U-bearing phases in weathering profiles with potential crystals, with the presence of epitaxial growths. These studies have also economic value (as in Nisa and Tarabau, where natural uranyl revealed natural heterogeneities affected by cationic substitution along phosphates of Cu and Ba were identified; Pinto et al., 2012; preferred crystallographic directions. The experimental results suggest Prazeres, 2011) and as fixing phases of U limiting its long-term, that the precipitation of metatorbernite is easier than metauranocircite. Simulations of the chemical system show that precipitation depends on million-year scale, dispersion in the oxidized surface supersaturation evolution, which in turn in a function of aqueous complex environment. It is thus important to understand the formation between phosphate and free uranyl ions. An electron probe thermodynamic stability of these phases and how they respond to microanalysis suggests that the failure to precipitate metauranocircite chemical changes of the surrounding environment. Building on may be due to later ionic depletion of the solution media (phosphate and previous knowledge on substitution mechanisms and uranyl), because of the early metatorbernite precipitation. crystallization relations between metatorbernite and Keywords: uranyl phosphates, uranium, metatorbernite, metauranocircite (Pinto et al., 2012; Sanchez-Pastor et al., 2013) metauranocircite, supergene environment. this work further investigates this relation in order to understand what drives the often late substitution of the Cu-phase by the Ba- Resumo: Os fosfatos de uranilo têm um papel determinante na phase, and how they compete for crystallization in the same imobilização do urânio em perfis de meteorização e enriquecimento chemical media. Hence, it’s of major importance to consider the supergénico. O presente trabalho incide na caracterização morfológica, general chemical formula of the autunite group estrutural e química destes minerais, naturais e sintéticos de Cu e/ou Ba – metatorbernite e metauranocircite. Análises em SEM revelaram uma (A(UO2)2(XO4)2·10-12H2O with A = Cu, Ca, Ba, or Mg and X = gama diversificada de morfologias, desde formas tabulares a rosáceas, P or As) in order to understand the cationic substitutions marcadas pela presença de crescimentos epitáxicos. Verificou-se também mentioned above. que os cristais naturais são heterogéneos, com substituições catiónicas ao longo de direcções cristalográficas preferenciais. Os resultados 2. Methods experimentais mostraram haver maior facilidade na precipitação de metatorbernite face à metauranocircite, mesmo nos sistemas com Cu e Ba The crystals studied were crystalized in laboratory in a silica-gel em solução. A modelação do sistema químico mostra que a precipitação é medium as described by Sánchez-Pastor et al. (2013). Also, função da evolução da sobressaturação, dependente da complexação do natural metatorbernite crystals from Musonoi Mine, Kolwezi, fosfato com o uranilo que baixa as respectivas actividades. Atendendo às análises de microssonda electrónica sugere-se que a precipitação de Katanga (Shaba) Province, Democratic Republic of Congo were metauranocircite seja inibida pelo empobrecimento em fosfato e uranilo immersed in silica gel and left to react with a BaCl2 solution. no sistema resultante da cristalização precoce de metatorbernite. Crystals were separated and cleaned for imaging and analysis Palavras-chave: fosfatos de uranilo, urânio, metatorbernite, metaurano- with an Environmental-SEM (HITACHI 3700N) working at 20 circite, ambiente supergénico. kV, a current ranging from 68 to 89 A, and low vacuum conditions (40 Pa) and equipped with a BRUKER Xflash 1Departamento de Geologia, Faculdade de Ciências da Universidade de 5010SDD EDS for qualitative chemical analyses. X-Ray micro- Lisboa, Edifício C6, Piso 4, Campo Grande, 1749-016 Lisboa, Portugal. diffraction of the crystals used a BRUKER Discovery X-ray 2 Instituto Dom Luiz, Faculdade de Ciências da Universidade de Lisboa, diffractometer with a Linxeye linear detector, a 0.3 mm Edifício C1, Piso 1, Campo Grande, 1749-016 Lisboa, Portugal. 3Departamento de Geociências e Laboratório HERCULES, Universidade de collimator, and a Göbel mirror. Radiation was generated from a Évora, Palácio do Vimioso, Largo Marquês de Marialva, 8, 7000-809 Évora, Cu-K lamp at 40 kV tension and 40 mA current. Scan ranged Portugal. from 3-75º 2, with 0.05º steps, and 2 s readings per step. Crystal *Autor correspondente/Corresponding author:[email protected] 28 Andrade et al. / Comunicações Geológicas (2015) 102, Especial I, 27-30 mounts were prepared for electron probe microanalysis (EPMA) limited Ba incorporation, as observed in SEM and EPMA (see and the crystals were analyzed for U, P, Cu, Ba, and Si (because below). of the gel medium). Astimex standards were used for Si The crystals from Cu-dominated experiments show (Diopside), Ba (Benitoite), U (UO2), P (Apatite), and Cu aggregates with tabular and micaceous habit, with sizes from 5 to (Cuprite). Analytical conditions used were a beam diameter of 20 10 m, also exhibiting macroscopic green colored dendritic m (reduced to 10m for the smaller crystals), 9 nA current (10 morphologies. Crystals from Ba-dominated experiments show nA for the smaller crystals), and 15 kV voltage. rosette-like morphologies, with sizes from 50 m to more than The porous silica-gel crystallizing environment was modelled 500 m, essentially yellow to brownish-white colored, in with PHREEQC (Parkhurst and Appelo, 1999) for transport and macroscopic acicular, capilar or dendritic aggregates. In the Cu multicomponent counter-diffusion of cationic and anionic species and Ba experiments, the crystals show similar morphologies to in the gel, without precipitation. The database used was the the previously described for both metatorbernite and phreeqc.dat with data on the solubility of metatorbernite and metauranocircite ranging from tabular to spheroidal, with sizes metauranocircite from Cretaz et al. (2013) and Vochten et al. from 50 to 100 m. They form macroscopic aggregates showing (1992), respectively. The U speciation was taken from the dendritic to tabular habits, along crystallization zones, where the llnl.dat. U diffusion coefficient was taken from Awakura et al. abundance and density of small crystals is significant. (1987). 3.2 Chemical features of the crystals 3. Results and Discussion The images showing EDS qualitative chemical maps revealed some particular aspects worth emphasizing: 3.1 Crystal morphologies 1. Epitaxial crystal growth is a mechanism of precipitation of Only 3 experiments were described by Sánchez-Pastor et al. a Ba-rich phase over metatorbernite (Fig. 2), which is shown by a (2013) which had a successful outcome, while the rest had higher signal of Ba over the topographic highs as opposed to a crystallization times > 2 years or did not crystallize anything. It lower signal elsewhere. was these ones that we used in this study, and their major 2. Natural metatorbernite crystals showed chemical difference to the set of experiments of Sánchez-Pastor et al. heterogeneities where a square domain core of metautunite (Ca) (2013) was that the gel in the horizontal column was reactive is surrounded by metatorbernite (Fig. 3), indicating that the (containing a fixed constant concentration of 50 ppm of U) cationic substitution in uranyl-phosphates of the metautunite instead of being diffused along with the remaining cations (Cu group affects a wider range of cations and mineral species than and/or Ba). Both experiment groups with either Cu or Ba in previously expected, as it has also been recently confirmed in solution revealed the presence of metatorbernite or natural environments (Prazeres et al., submitted). metauranocircite, respectively, showing different morphologies 3. Cation substitutions of both Ca2+ and Cu2+ for Ba2+ in and crystal aggregates. However, experiments with Cu and Ba in natural samples (Fig. 4). This substitution occurs pervasively solution showed only metatorbernite crystals with no direct along structural discontinuities, essentially through cleavage indication of the presence of a Ba-phase (Fig. 1).This was only plans and other crystallographic directions with a clear detected in some crystals as mixed phases of Cu and Ba with progression inward the crystal.

Fig. 1. SEM images for the different crystal morphologies: (a) metauranocircite; (b) metatorbernite; (c) to (f) metatorbernite crystals in the Cu and Ba system. Fig. 1. Imagens de microscópio electrónico de varrimento evidenciando as diferentes morfologias: (a) metauranocircite; (b) metatorbernite; (c) a (f) cristais de metatorbernite em sistemas de Cu e Ba. Cationic substitution in uranyl phosphates 29

Fig. 4. SEM images (EDS map/BSE) of Ca substitution with Ba along cleavage plans and grain boundaries. Fig. 4. Imagens obtidas por MEV (mapa EDS/BSE) mostrando a substituição de Ca com Ba ao longo dos planos de clivagem e nas fronteiras de grãos.

Fig. 2. SEM image of secondary electrons superposed with the Ba signal showing what may be crystal epitaxial growths over metatorbernite. Fig. 2. Imagens de eletrões secundários de MEV sobrepostos com o sinal de Ba, mostrando possíveis crescimentos epitáxicos sobre a metatorbernite.

Fig. 5. Compositional EPMA map (a) showing the variation in Cu and Ba content within the same crystal and the correspondent zoomed mineral SEM image (b). Fig. 5. Mapa composicional obtido por Microanálise em Sonda Electrónica (EPMA) mostrando (a) a variação no conteúdo de Cu e Ba num mesmo cristal e (b) Imagem SEM com a ampliação do respetivo mineral.

These longer experimental times (1000 h) gradually stabilize the chemical system into a steady-state in which the system does not evolve but clearly shows the domains where supersaturation conditions are maintained. The rest is the result of the kinetics of crystallization and crystal growth, in which to obtain sizeable crystal dimensions to work with at least 2 years were necessary to grow the uranyl-phosphate crystals. Experiment R2 (with Cu and Ba as diffusing cations) shows that the nucleation front migrates towards the left as phosphate diffuses from the right and complexes with uranium in the silica gel (Fig. 6a). Modelling Fig. 3. SEM image of secondary electrons with superimposed signals for Ca and Cu results also show the key role of uranyl in solution (U10 – with showing the heterogeneous character of the metatorbernite crystals. Cu and Ba as diffusing cations), because in this experiment (Fig. Fig. 3. Imagens de eletrões secundários de MEV sobrepostos com o sinal de Ba e Cu 6b) as in some others, crystals were either scant or failed to mostrando o caráter heterogéneo dos cristais de metatorbernite. precipitate altogether. In the initial stage, and because we used reactive gel with U, phosphate diffusion establishes a low EPMA on the synthetic crystals confirm the previous concentration profile and limited complexation with U allowing observations on SEM, admitting the existence of mixed crystals supersaturation conditions in the middle part of the tube allowing (Fig. 5) with a variable proportion of Cu:Ba from 0.13:0.97 to the precipitation of metauranocircite and/or metatorbernite. 0.37:0.85. Furthermore, the analyses didn’t reflect an exact 1:1 2+ However, as the system evolves in time, the diffusion of stoichiometry of U:P, but rather richer in uranium (UO2 ) 3- phosphate raises its concentration along the tube which in turn relative to phosphorous (PO4 ). This can be either the result of 2+ 2+ complexes the U in the gel, reducing their activity in solution and cation substitution in the interlayers (Ba for Cu ) or the depressing the SI so much that further nucleation of crystals in presence of hypothetical vacancies. Nevertheless, Pinto et al. the gel is prevented. (2012) already observed such stoichiometries in natural Therefore, only in a very limited time window and chemical metatorbenite-metauranorcicite crystals. conditions does uranyl-phosphate minerals precipitate, unless U concentration in the gel is higher as in the other experiments. 3.3 Modelling the chemical system Geochemical modelling simulations show the distribution, along 4. Conclusions the tubes, of species concentration and the saturation index (SI) The most relevant observation in this work is the preferential of metatorbernite and metauranocircite. The location of the precipitation of metatorbernite relative to metauranocircite in highest SI should coincide with the highest rates of crystal conditions apparently favourable for both minerals to nucleate precipitation and growth in the tubes. While short time and grow. This is somewhat counterintuitive since the processes simulations (up to 100 h) showed results that were not coherent of Cu (and Ca) substitution by Ba are common in nature (Pinto et with experimental observations, extended simulation, up to 1000 al., 2012; Prazeres et al., submitted) and in experiments. This h, showed areas of higher growth rates coincident with what is difference may be primarily due to kinetic factors, but is at experimentally observed (real experiments lasted more than 2 present unresolved. years, actually). 30 Andrade et al. / Comunicações Geológicas (2015) 102, Especial I, 27-30

(a) (b)

Fig. 6. PHREEQC modelling results for R2 (a) and U10 (b) experiments. Explanation in text. Fig. 6. Resultados da modelação em PHREEQC para as experiências R2 (a) e U10 (b). Explicação no texto.

The metatorbernite crystals obtained in a Ba and Cu solution environment also have bigger dimensions than those in References monocationic systems. Awakura, Y., Sato, K., Majima, H., Hirono, S., 1987. The measurement The reacted natural samples with BaCl2 solutions showed that secondary cationic substitution of Ca and Cu by Ba is of the diffusion coefficient of U(VI) in aqueous uranyl sulfate thermodynamically favourable. However, epitaxial growth may solutions. Metallurgical Transactions, 18B. Cretaz, F., Szenknect, S., Clavier, N., Vitorge, P., Mesbah, A., Descoste, be the process by which Ba-rich phases preferentially grow M., Poinssot, C., Dacheux, N., 2013. Solubility properties of synthetic which also explains the variety of forms displayed by and natural meta-torbernite. Journal of Nuclear Materials, 442: 195- metatorbernite crystals precipitated in mixed chemical 207. environments (Sanchez-Pastor et al., 2013). Chemical modelling Parkhurst, D. L. and Appelo, C. A. J., 1999. User's guide to PHREEQC showed that mineral precipitation depends on how (version 2) - A computer program for speciation, batch-reaction, one- supersaturation evolves as a function of the free chemical dimensional transport, and inverse geochemical calculations. U.S. species, especially with the complexation between uranyl and Geological Survey Water-Resources Investigations Report, 99-4259, phosphate. The absence and low abundance of metauranocircite 312. Pinto, A. J., Gonçalves, M. A., Prazeres, C., Astilleros, J. M., Batista, M. in the experiments may be related to the solubility difference J., 2012. Mineral replacement reactions in naturally occurring relative to metatorbernite, but we must bear in mind that accurate hydrated uranyl phosphates from the Tarabau deposit: Examples in the and sufficiently reliable solubility data for these minerals is Cu-Ba uranyl phosphate system. Chemical Geology, (312-313): 18-26. lacking. EPMA results show only evidence for the joint Prazeres, C. M., 2011. Caracterização geoquímica, radiométrica e metauranocircite and metatorbernite crystallization in the form of mineralógica de algumas mineralizações de urânio da região de Nisa. mixes crystals, mostly copper-rich phases with a limited MSc thesis, Universidade de Lisboa, 147. incorporation of barium. This result implies that Ba should build- Prazeres, C., Batista, M. J., Pinto, A., Gonçalves, M. A. (submitted). up its concentration in the gel with time, but still conditions are Geochemical environment and radiometric mapping of a vein-related uranium deposit: insights into uranium distribution and mobility in the not favourable for the Ba-phase to precipitate. Therefore, another weathering zone of the Nisa deposit, Portugal. Submitted to Journal of explanation is that later nucleation may be hindered by the Geochemical Exploration. depletion of ions (uranyl and phosphate) in solution, as shown by Sánchez-Pastor, N., Pinto, A. J., Astilleros, J. M., Fernández-Diáz, L., the numeric models. Gonçalves, M. A., 2013. Raman spectroscopic characterization of asynthetic, non-stoichiometric Cu-Ba uranyl phosphate. Spectrochimica Acta Part A: Molecular and Biomolecular Acknowledgments Spectroscopy 113: 196-202. André Pinto is thanked for providing the reacted metatorbernite Vochten, R. F., Haverbeke, L. Van, Springel, K. Van., 1992. crystals with the Ba solutions in gel. Transformation of chernikovite into meta-uranocircite II, Ba(UO2)2(PO4).6H2O and study of its solubility. Mineralogical Magazine, 56: 367-372.