Monazite Alteration in H 2 O ̆ HCl ̆ NaCl ̆ CaCl 2 Fluids at 150 ̋ C and p sat : Implications for Uranium Deposits

Spectacular alteration of monazite by diagenetic/hydrothermal brines is well documented in some Proterozoic sedimentary basins in close relationship with high-grade uranium (U) deposits. Hence, monazite has been proposed as a viable source for some U deposits. However, monazite alteration remains enigmatic with regard to its high stability in relatively low temperature hydrothermal conditions. Here, the results of batch experiments in which 10 mg of natural monazite grains were reacted with 15 mL of Na-Ca-Cl (6 molal Cl) solutions as well as in pure water at 150  ̋C and saturated vapor pressure (psat) for one and six months are reported. The influence of pH (pH = 1, 3, 7) and relative molar proportions of Na and Ca (Na/(Na + Ca) = 0, 0.5, 1), were tested. Discrete alteration features (etch pits and roughened surfaces) appear in a minority of the one month experiments and are more developed in the six months experiments, especially at pH = 1 and 3. Although spectacular alteration of monazite, as seen around U deposits, could not be reproduced here, this study shows that monazite is unstable in the presence of fluids analogous to acidic deep basinal brines.

Spectacular alteration of monazite in the presence of deep basinal brines is well documented in some Proterozoic sedimentary basins and their crystalline basements in close relationship with high-grade U deposits, especially in the Athabasca Basin (Canada), McArthur Basin (Australia) and Franceville Basin (Gabon) [16][17][18]39,40].The aim of this work is to test experimentally the stability of monazite in the conditions that have prevailed during the formation of world-class U deposits in the Athabasca, McArthur and Franceville Basins.For this purpose, batch experiments have been carried out in which natural monazite grains were reacted with 6 molal Cl, Na-Ca-Cl brines with variable Na/(Na + Ca) molar ratios and pH at 150 ˝C and p sat in autoclaves with one and six months run times.

Background
The geological background of the Proterozoic basins where spectacular alteration of monazite has been described, the world-class uranium deposits within those basins, the petrography of altered monazites and the fluid conditions that have prevailed during alteration of monazite are reported here.

Proterozoic Basins and Uranium Deposits
High-grade and large-tonnage UO 2 deposits (up to 200 kt of U at 20% U on average at McArthur River deposit) are located close to the basement unconformity of the Paleoproterozoic Athabasca Basin (Saskatchewan, Canada) [41,42].These U deposits originate from protracted basin-scale circulation of U-bearing fluids, at ca. 1.5-1.2Ga, through sedimentary and basement rocks concurrent with sediment diagenesis at temperatures of 150-200 ˝C [43][44][45][46][47].This style of "unconformity-related" uranium deposits is unique to a restricted number of Proterozoic basins including the Athabasca and Thelon Basins (Canada) [48,49] and the McArthur Basin (Northern Australia) [50][51][52].Comparable processes may have been responsible for the famous Oklo U deposits in the Franceville Basin (Gabon), which on account of their high grades and greater age "2 Ga formed natural nuclear reactors [46].

Monazite Alteration and U Sources in Proterozoic Basins and Basements
Potential sources of uranium for U deposits in those Proterozoic basins include mainly: (i) accessory minerals (apatite, zircon, monazite) in the basins and their basement [16,17,39,40,44,53] and (ii) vein-type UO 2 within metamorphic rocks and magmatic UO 2 within granitic pegmatites, as described in the basement rocks of the eastern margin of the Athabasca Basin [54][55][56].
Minerals 2015, 5, page-page modifications (e.g., Na-Ca, Na-K, Ca-Mg exchanges, U leaching) after interacting with basin and basement rocks as seen from the differences between evaporated seawater and fluid inclusions compositions [46,61,63,65].Examples of intensive alteration of monazite by diagenetic/hydrothermal brines from the basement of the Proterozoic Athabasca Basin (Canada): (a) Back-scattered electron image of a monazite grain (grey), which is nearly completely altered to Th-silicate (white) (modified from [17]).(b) Back-scattered electron image of a relict monazite (white) associated with aluminum phosphate-sulfate mineral (APS) (grey) (modified from [57]).

Experimental
The preparation of the starting material for the experiments (natural monazite from Manangotry, Madagascar), the composition of the starting material, the preparation of the experiments, the experimental conditions and the analytical conditions for scanning electron microscope analyses of monazite grains before and after the experiments, are described hereby.

Scanning Electron Microscope Analyses
Secondary electron and backscattered electron imaging of monazite grains (before and after experiments) was performed using a JEOL JSM7600F scanning electron microscope (SEM) (JEOL Ltd., Tokyo, Japan) equipped with energy dispersive spectrometer (EDS) and a wavelength dispersive spectrometer (WDS) at the Service Commun de Microscopies Electroniques et de Microanalyses (SCMEM, GeoRessources lab., Nancy, France).
Examples of intensive alteration of monazite by diagenetic/hydrothermal brines from the basement of the Proterozoic Athabasca Basin (Canada): (a) Back-scattered electron image of a monazite grain (grey), which is nearly completely altered to Th-silicate (white) (modified from [17]).(b) Back-scattered electron image of a relict monazite (white) associated with aluminum phosphate-sulfate mineral (APS) (grey) (modified from [57]).

Experimental
The preparation of the starting material for the experiments (natural monazite from Manangotry, Madagascar), the composition of the starting material, the preparation of the experiments, the experimental conditions and the analytical conditions for scanning electron microscope analyses of monazite grains before and after the experiments, are described hereby.

Scanning Electron Microscope Analyses
Secondary electron and backscattered electron imaging of monazite grains (before and after experiments) was performed using a JEOL JSM7600F scanning electron microscope (SEM) (JEOL Ltd., Tokyo, Japan) equipped with energy dispersive spectrometer (EDS) and a wavelength dispersive spectrometer (WDS) at the Service Commun de Microscopies Electroniques et de Microanalyses (SCMEM, GeoRessources lab., Nancy, France).

Sample Preparation
The monazite monocrystal was first crushed in a confined hydraulic press.The crushed material was then sieved and the >200 µm fraction was hand-crushed in an agate pestle and mortar.All the crushed material was sieved again and the 60-105 µm fraction was recovered for the experiments.The 60-105 µm fraction of crushed material was processed by a Franz magnetic separator in order to remove mineral impurities (plagioclase, pyroxene, apatite, and thorite).Optical examination under binoculars of the resulting 60-105 µm fraction revealed <1% mineral impurities (Figure 2b).The 60-105 µm fraction was then rinsed twice in distilled water.Examination of crushed monazite grains by secondary electron imaging shows clean surfaces and angular grain shapes (Figure 2c).No indication of metamictization has been observed.Examination of crushed and mounted monazite grains by back-scattered electron imaging shows highly homogeneous composition (Figure 2d), as seen by electron microprobe analysis (Table 1).

Sample Preparation
The monazite monocrystal was first crushed in a confined hydraulic press.The crushed material was then sieved and the >200 µm fraction was hand-crushed in an agate pestle and mortar.All the crushed material was sieved again and the 60-105 µm fraction was recovered for the experiments.The 60-105 µm fraction of crushed material was processed by a Franz magnetic separator in order to remove mineral impurities (plagioclase, pyroxene, apatite, and thorite).Optical examination under binoculars of the resulting 60-105 µm fraction revealed <1% mineral impurities (Figure 2b).The 60-105 µm fraction was then rinsed twice in distilled water.Examination of crushed monazite grains by secondary electron imaging shows clean surfaces and angular grain shapes (Figure 2c).No indication of metamictization has been observed.Examination of crushed and mounted monazite grains by back-scattered electron imaging shows highly homogeneous composition (Figure 2d), as seen by electron microprobe analysis (Table 1).

Batch Experiments
Experimental solutions were prepared using (i) distilled water with a resistance over length of 18.2 MΩ•cm −1 ; (ii) CaCl2-2H2O PROLABO 22322.295powder; (iii) NaCl AnalR NORMAPUR 95%-97% powder; and (iv) hydrochloric acid provided by Riedel de Haen as HCl-H2O mixture saturated with HCl (37% by mass).A total of twelve experimental solutions were prepared.Three solutions were prepared with variable Na/(Na + Ca) molar ratios of 0, 0.5 and 1.Then, from each of the three solutions with different Na/(Na + Ca) values, three different solutions were prepared with pH values of 1, 3 and 7, adjusted with HCl and a pH-meter calibrated using calibration buffer solutions.All of those nine solutions have a total Cl concentration of 6 molal.Finally, three solutions were prepared with distilled water and no salts and pH values of 1, 3 and 7, adjusted with HCl and a pH-meter.For each experiment, 10 mg of purified 60-105 µm fraction of monazite grains were weighted in a Mettler AT 201 (Mettler-Toledo, Greifensee, Switzerland) precision balance.Monazite

Batch Experiments
Experimental solutions were prepared using (i) distilled water with a resistance over length of 18.2 MΩ¨cm ´1; (ii) CaCl 2 -2H 2 O PROLABO 22322.295powder; (iii) NaCl AnalR NORMAPUR 95%-97% powder; and (iv) hydrochloric acid provided by Riedel de Haen as HCl-H 2 O mixture saturated with HCl (37% by mass).A total of twelve experimental solutions were prepared.Three solutions were prepared with variable Na/(Na + Ca) molar ratios of 0, 0.5 and 1.Then, from each of the three solutions with different Na/(Na + Ca) values, three different solutions were prepared with pH values of 1, 3 and 7, adjusted with HCl and a pH-meter calibrated using calibration buffer solutions.All of those nine solutions have a total Cl concentration of 6 molal.Finally, three solutions were prepared with distilled water and no salts and pH values of 1, 3 and 7, adjusted with HCl and a pH-meter.For each experiment, 10 mg of purified 60-105 µm fraction of monazite grains were weighted in a Mettler AT 201 (Mettler-Toledo, Greifensee, Switzerland) precision balance.Monazite grains and 15 mL of experimental solutions were then placed in twelve 20 mL polytetrafluoroethylene (PTFE) containers, which were in turn placed in twelve autoclaves, and heated at 150 ˝C under saturated vapor pressure (p sat ) in a sealed ceramic oven.After one month and six months, each autoclave was extracted, quickly cooled by water quenching, and opened.

Results
Experimental conditions, experiment numbers and observations by SEM of monazite grains after one month and six months experiments are reported in Tables 2 and 3 respectively.Observations are reported hereafter according to experiment numbers (EXP#).SEM images of the different alteration features on monazite grains are shown in Figures 3 and 4 for one month and six months experiments respectively.All experiments at pH = 7 (EXP# 3, 6, 9, 12) and experiments with CaCl 2 only (EXP# 1, 2, 3) show no alteration of the surface of monazite grains.Only EXP# 4, 5 and 7 show some etch pits (Figure 3a,b) and/or roughened surfaces (Figure 3c,d).Note that those alteration features were not systematically observed on all grain surfaces for a given experiment.This could be due to the fact that some surfaces were not or very little exposed to the experimental solutions at places where grains were close to each other.

Six Months Experiments
Six months experiments show significantly more developed alteration features compared to the one month experiments, although not widespread.Types of alteration features are similar to those in one month experiments and consist of etch pits and roughened surfaces.Etch pits seem to develop first, leading to roughened surfaces when the pit density in high (Figure 4a-e).Similarly to one month experiments, alteration features were not systematically observed on all grain surfaces for a given experiment, possibly due to the fact that some surfaces were not or very little exposed to the experimental solutions at places where grains were close to each other.Etch pits and roughened surfaces are systematically visible in experiments at pH = 1 (EXP# 13,16,19,22).All experiments at pH = 7 (EXP# 15,18,21,24) show no alteration of the surface of monazite grains.The most intensively altered monazite grains form in EXP# 17 where roughened surfaces are well developed (Figure 4e).Roughened surfaces appear to be heterogeneous in composition; however their small-scale compositional variability does not permit further investigation (Figure 4f).

Six Months Experiments
Six months experiments show significantly more developed alteration features compared to the one month experiments, although not widespread.Types of alteration features are similar to those in one month experiments and consist of etch pits and roughened surfaces.Etch pits seem to develop first, leading to roughened surfaces when the pit density in high (Figure 4a-e).Similarly to one month experiments, alteration features were not systematically observed on all grain surfaces for a given experiment, possibly due to the fact that some surfaces were not or very little exposed to the experimental solutions at places where grains were close to each other.Etch pits and roughened surfaces are systematically visible in experiments at pH = 1 (EXP# 13,16,19,22).
All experiments at pH = 7 (EXP# 15,18,21,24) show no alteration of the surface of monazite grains.The most intensively altered monazite grains form in EXP# 17 where roughened surfaces are well developed (Figure 4e).Roughened surfaces appear to be heterogeneous in composition; however their small-scale compositional variability does not permit further investigation (Figure 4f).
Minerals 2015, 5, page-page 7 H2O-NaCl at 800 °C [72].The etch pits and roughened surfaces show some morphological similarities with "leached surfaces" observed for experiments in aqueous solutions at 100-250 °C and psat [73] and dissolution near ledge features observed for experiments in 20-230 °C acidic aqueous solutions [71].Therefore, etch pits combined with the absence of residual crystals on grain surfaces may indicate congruent monazite dissolution.Conversely, the heterogeneous nature of roughened surfaces (Figure 4f) is more indicative of fluid-mediated dissolution-reprecipitation.

Discussion
The influence of the different experimental parameters and the implications of the experiments results on monazite stability in H 2 O ˘HCl ˘NaCl ˘CaCl 2 fluids at 150 ˝C and p sat for uranium deposits, are discussed here.

Significance of Alteration Features
Etch pits observed here look similar to those observed for experiments involving brines at 500 ˝C and 500 MPa [24] and dissolution pits, grooves and patterns observed for experiments in H 2 O-NaCl at 800 ˝C [72].The etch pits and roughened surfaces show some morphological similarities with "leached surfaces" observed for experiments in aqueous solutions at 100-250 ˝C and p sat [73] and dissolution near ledge features observed for experiments in 20-230 ˝C acidic aqueous solutions [71].
Therefore, etch pits combined with the absence of residual crystals on grain surfaces may indicate congruent monazite dissolution.Conversely, the heterogeneous nature of roughened surfaces (Figure 4f) is more indicative of fluid-mediated dissolution-reprecipitation.

Influence of Starting Material
The experimental setup used here is very simple and does not include additional minerals besides monazite.Minerals unstable in those conditions could promote alteration of monazite by supplying ions necessary to form secondary phases as observed in nature like Si, Al or S to form APS and Th-silicates.From the few published experiments at relatively low temperature, monazite replacement by REE-rich steacyite under P-T conditions of 250-350 ˝C and 200-400 MPa, with starting materials including monazite + albite + K-feldspar + muscovite + biotite + SiO 2 + CaF 2 and Na 2 Si 2 O 5 + H 2 O has been demonstrated [32].Similarly, experiments involving monazite + albite ˘K-feldspar + muscovite ˘biotite + SiO 2 + CaF 2 and a variety of fluids, at 450 to 500 ˝C and 450 to 610 MPa, showed alteration of monazite to allanite, REE-epidote, fluorapatite, and/or fluorapatite-britholite [24].Further experiments using for example granitic powder in order to mimic the mineralogical environment of monazite in the basement of the Athabasca and McArthur Basins could help to better approaching natural conditions for monazite alteration in U deposits from Proterozoic basins.

Influence of Temperature and Pressure
One possible limiting factor explaining that alteration of monazite as seen in Proterozoic basins could not be reproduced in the present experiments could be the relatively low temperature of 150 ˝C.Although temperature may have evolved during the lifespan of the hydrothermal systems in the Proterozoic Athabasca, McArthur and Franceville Basins, with various estimates ranging from 100 to 250 ˝C [40,45,50,63], most fluid inclusion data point to 150 ˝C as a reasonably good representative temperature for the NaCl-rich and CaCl 2 -rich brines.Monazite alteration and uranium mobilization by Na-Ca-Cl brines at temperature >500 ˝C have been demonstrated in some uranium deposits in the Western Zambian Copperbelt [74,75].Even though such high temperature is unlikely to have been reached in the Athabasca, McArthur and Franceville Basins, it cannot be excluded that relatively short high temperature events, significantly hotter than the average temperature of 150 ˝C, have occurred and could be associated with monazite alteration events.Factors such as fluid density and dynamic viscosity will be significantly affected by the pressure prevailing during the experiments.Here, p sat is well below the pressure of ca.0.5 to 1.2 kbars at the base of the Proterozoic Basins [40,45,50,63].However, from the current knowledge on monazite solubility, the influence of pressure on monazite alteration in the present experimental conditions is hardly predictable [76].

Influence of Experiment Duration
Although the six month experiments do not show intense alteration of monazite, they produced significantly more alteration features than the one month experiments.This shows that monazite destabilization in those conditions is possible but is kinetically probably a very slow process.The duration of the hydrothermal systems for the formation of the Proterozoic uranium deposits in the Athabasca, McArthur and Franceville Basins is poorly constrained but various estimates are in the order of 0.1-100 Myr years [47,77,78].It is therefore clear that experimental conditions applied here can hardly approach the duration of those hydrothermal systems.The total amount of monazite dissolved in the present experiments is hard to constrain, most probably below 0.1%, possibly in the order of 0.001% to 0.01%.If between 0.001% and 0.01% of the monazite is dissolved in the six months experiments, it would theoretically take 5 ˆ10 3 to 5 ˆ10 4 years to dissolve all 10 mg of monazite at a constant rate.Therefore, complete destabilization of monazite may have required protracted fluid flow.However, the duration for complete destabilization of monazite was probably below the lower estimate for the duration of the hydrothermal systems at the origin of uranium deposits in Proterozoic basins.

Influence of Fluid/Rock Ratio
With 10 mg of monazite grains and 15 mL of experimental solutions, the mass fluid/rock ratio in the experiments is 1.5 ˆ10 3 .This is relatively high for a usual hydrothermal system but below the fluid/rock ratios of >10 4 calculated after chemical mass balance for some breccia systems associated to uranium deposits in the Athabasca Basin [79].Intense alteration of monazite in the uranium deposits in the Proterozoic basins is therefore a possible consequence of those exceptionally high fluid/rock ratios which, according to the fluid flow velocity estimates in those environments [77], agrees with durations of the hydrothermal systems in the order of 0.1-100 Myr years.Further experiments with, for instance, 1 mg of monazite grains could help test this hypothesis.However, in nature, the fluids were mobile and not static and equilibrium between the solid and the fluid was probably not achieved, which makes comparison of fluid/rock ratios between nature and the present static experiments difficult.

Influence of pH and Fluid Composition
The experimental results show a clear influence of pH in the destabilization of monazite, with increasing abundance of alteration features from pH = 7 to pH = 1.This is consistent with the acidic nature of the brines in the Athabasca Basin, which is also a prerequisite for efficient uranium transport [47].Experiments in pure water with variable pH show no alteration features except for pH = 1 during six months experiments.Therefore, pH and not the ionic strength of the brines, is the dominant control on monazite alteration under these conditions.The Na/(Na + Ca) ratio appears to have also some control on monazite destabilization with increasing abundance of alteration features from Na/(Na + Ca) values of 0 to 1 and then 0.5.CaCl 2 -rich brines have been shown to carry ca.one order of magnitude more uranium that NaCl-rich brines in the Athabasca Basin [61,63].If monazite is effectively the main uranium source in those brines, this could mean that brines with a low Na/(Na + Ca) ratio would be more aggressive with respect to monazite.This would also be consistent with the observed formation of Ca-Th phosphates and Ca-bearing aluminum phosphate-sulfate minerals as a product of monazite destabilization in the Athabasca Basin [17].However, this hypothesis cannot be confirmed by the present results.Instead, the results indicate that mix-salt solutions are the most efficient for monazite destabilization.This requires a mixing of the NaCl-rich and CaCl 2 -rich brine end-members prior to destabilizing the monazite.It is noteworthy that the alteration features are best developed for pH = 3 and a Na/(Na +Ca) ratio of 0.5 in the six months experiments.Those conditions appear to be the closest to the natural conditions where pH is thought to be between 3 and 4, and brine mixing is widespread [45][46][47]61,63].Finally, further experiments could be carried out with fluid composition approaching more closely the compositions of the NaCl-rich and CaCl 2 -rich brines, as determined for the Athabasca Basin by [61,63].Indeed, the composition of the NaCl-rich and CaCl 2 -rich brines in the Athabasca Basin, with notably high Mg 2`a nd K `content besides Na `and Ca 2`, is fairly unusual compared to more usual diagenetic brines largely dominated by Na-Ca-Cl [80].Further experiments with KCl and MgCl 2 -rich brines and mixed-salt brines could help better understanding the competing roles of cations for monazite destabilization in those solutions and possibly better constrain which of the so-called "NaCl-rich brine" or "CaCl 2 -rich brine" or their mixing terms is responsible for the spectacular alteration of monazite in U deposits in Proterozoic basins.

Conclusions
Taken together, batch experiments using 10 mg of natural monazite grains and 15 mL of acidic to neutral Na-Ca-Cl (6 molal Cl) brines at 150 ˝C and p sat for one month and six months show limited alteration of the monazite grains.Some discrete alteration features (etch pits and roughened surfaces) have developed, especially in acidic solution, indicating that natural monazite is unstable in these conditions, but alteration is very slow.This contrasts with spectacular alteration of monazite in analogous natural conditions observed in the Proterozoic Athabasca, McArthur and Franceville Basins where monazite is partially to fully replaced by Th-silicates and aluminum phosphate-sulfate (APS) minerals.Spectacular alterations observed in nature may therefore have required more protracted fluid/rock interaction, higher fluid/rock ratios or unusual brine composition.Further experiments involving additional reactive minerals together with monazite (e.g., granite powder) could also provide more favorable conditions for monazite alteration, closer to natural conditions.

Figure 1 .
Figure 1.Examples of intensive alteration of monazite by diagenetic/hydrothermal brines from the basement of the Proterozoic Athabasca Basin (Canada): (a) Back-scattered electron image of a monazite grain (grey), which is nearly completely altered to Th-silicate (white) (modified from[17]).(b) Back-scattered electron image of a relict monazite (white) associated with aluminum phosphate-sulfate mineral (APS) (grey) (modified from[57]).

Figure 2 .
Figure 2. Starting experimental material: (a) Photograph of the natural monazite monocrystal from Manangotry (Madagascar) selected for this study (sample PK6 in [70]).(b) Photograph under binoculars of crushed, sieved and purified by magnetic separation, 60-105 µm monazite grains from the monocrystal shown in (a).(c) Secondary electron image of one of the monazite grains shown in (b).(d) Back-scattered electron image of some of the mounted monazite grains shown in (b).

Figure 2 .
Figure 2. Starting experimental material: (a) Photograph of the natural monazite monocrystal from Manangotry (Madagascar) selected for this study (sample PK6 in [70]).(b) Photograph under binoculars of crushed, sieved and purified by magnetic separation, 60-105 µm monazite grains from the monocrystal shown in (a).(c) Secondary electron image of one of the monazite grains shown in (b).(d) Back-scattered electron image of some of the mounted monazite grains shown in (b).

Figure 3 .
Figure 3. SEM images of the different alteration features on monazite grains after one month experiments: (a) etch pits in EXP# 7; (b) etch pits in EXP# 5; and (c,d) roughened surfaces in EXP# 4.

Figure 4 .
Figure 4. SEM images of the different alteration features on monazite grains after the six months experiments: (a) etch pits in EXP# 20; (b) etch pits and roughened surface in EXP# 13; (c) etch pits and roughened surface in EXP# 22; (d) etch pits and roughened surface in EXP# 16; (e) roughened surface in EXP# 17; and (f) zoom into roughened surface in EXP# 17.

Figure 4 .
Figure 4. SEM images of the different alteration features on monazite grains after the six months experiments: (a) etch pits in EXP# 20; (b) etch pits and roughened surface in EXP# 13; (c) etch pits and roughened surface in EXP# 22; (d) etch pits and roughened surface in EXP# 16; (e) roughened surface in EXP# 17; and (f) zoom into roughened surface in EXP# 17.

Table 2 .
Experimental conditions, experiment numbers (#) and observations by SEM of monazite grains after one month experiments.

Table 3 .
Experimental conditions, experiment numbers and observations by SEM of monazite grains after six months experiments.One month experiments with pure H 2 O and H 2 O-HCl mixtures (pH = 1, 3, 7) (EXP# 10, 11, 12) show no alteration of the surface of monazite grains.Experiments with NaCl and/or CaCl 2 show contrasting results.EXP# 1, 2, 3, 6, 8, 9 show no alteration of the surface of monazite grains.