Subscriber access provided by University of Sussex Library
Article
Recovery of Rare Earth Elements and Yttrium from Passive Remediation Systems of Acid Mine Drainage Carlos Ayora, Francisco Macías, Ester Torres, Alba Lozano, Sergio Carrero, Jose Miguel Nieto, Rafael Perez-Lopez, Alejandro Fernandez-Martinez, and Hiram A. Castillo-Michel Environ. Sci. Technol., Just Accepted Manuscript • DOI: 10.1021/acs.est.6b02084 • Publication Date (Web): 28 Jun 2016 Downloaded from http://pubs.acs.org on July 2, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Environmental Science & Technology is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 33
Environmental Science & Technology
1 2
Recovery of Rare Earth Elements and Yttrium from Passive Remediation Systems of Acid Mine Drainage
3
Carlos Ayora (*,1), Francisco Macías (2), Ester Torres (1), Alba Lozano (1), Sergio Carrero (2),
4
José-Miguel Nieto (2), Rafael Pérez-López (2), Alejandro Fernández-Martínez (3) and Hiram
5
Castillo-Michel (4)
6
(1) Institute of Environmental Assessment and Water Research, CSIC, c/ Jordi Girona 18,
7
08034
Barcelona,
Spain,
[email protected],
[email protected],
8
[email protected], tel. 34-934006100, fax. 34-932045904
9
(2) Department of Geology, University of Huelva, Campus ‘El Carmen’ s/n
10
21071 Huelva, Spain,
[email protected],
[email protected],
11
[email protected],
[email protected], tel. 34-959219809, fax. 34-
12
959219810
13
(3) Institute de Sciences de la Terre, CNRS & Univ. Grenoble Alpes, 1381 Rue de la Piscine,
14
38041 Grenoble, France,
[email protected], tel. 33-
15
476635197, fax. 33-476635201
16 17 18
(4) European Synchrotron Radiation Facility, 6 rue Jules Horowitz, Grenoble, France,
[email protected], tel. 33-476882948, fax. 33-476882785 (*) Corresponding author
19 20
Abstract
21
Rare Earth Elements and Yttrium (REY) are raw materials of increasing importance for modern
22
technologies, and finding new sources has become a pressing need. Acid Mine Drainage (AMD)
23
is commonly considered an environmental pollution issue. However, REY concentrations in
24
AMD can be several orders of magnitude higher than in naturally occurring water bodies. With 1 ACS Paragon Plus Environment
Environmental Science & Technology
Page 2 of 33
25
respect to shale standards, the REY distribution pattern in AMD is enriched in intermediate and
26
valuable REY, such as Tb and Dy. The objective of the present work is to study the behavior of
27
REY in AMD passive remediation systems. Traditional AMD passive remediation systems are
28
based on the reaction of AMD with calcite-based permeable substrates followed by
29
decantation ponds. Two column experiments simulating AMD treatment demonstrate that
30
schwertmannite does not accumulate REY, which, instead, are retained in the basaluminite
31
residue. The same observation is made in two field-scale treatments from the Iberian Pyrite
32
Belt (IPB, SW Spain). Based on the amplitude of this process, and on the extent of the IPB, our
33
findings suggest that the proposed AMD remediation process can represent a modest but
34
suitable REY source. In this sense, the IPB could function as a giant heap leaching process of
35
regional scale, in which rain and oxygen act as natural driving forces with no energy
36
investment. In addition to the environmental benefits of its treatment, AMD is expected to last
37
for hundreds of years, and therefore the total reserves are practically unlimited.
38
Key-words: acid rock leaching, schwertmannite, basaluminite, fluorite, Iberian Pyrite Belt,
39
sustainable mining, renewable resource.
40
2 ACS Paragon Plus Environment
Page 3 of 33
Environmental Science & Technology
41 42
Table of Contents (TOC) Art AMD passive remediation treatment
43
AMD
pH3.5 Fe Sulfide oxidation and rock leaching
REY ORE
pH>5.5 Al limestone
Treated water
3 ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 33
44
INTRODUCTION
45
Rare Earth Elements (REE) together with Yttrium (REY) are essential raw materials for modern
46
technological developments. Their most important uses include the manufacturing of
47
permanent magnets for wind turbines, alloys for rechargeable batteries and jet engines, and
48
phosphor light-emitting compounds for plasma, liquid crystal or light-emitting diodes. In 2011,
49
global demand was 105 kt of REY oxides, and it is expected to grow to 160 kt by 2016.1 In
50
general, global consumption of REY is expected to increase at a compound annual growth rate
51
in excess of 5% from 2014 through 2020.2 This increasing demand is particularly evident for
52
elements used in wind energy and electric vehicles, such as Dy and Nd. In the absence of
53
drastic changes in the present-day technologies of reuse and recycling, increases of 700% and
54
2,600% for Nd and Dy, respectively, are expected over the next 25 years.3
55
Most mined REY deposits are located in carbonatites and other alkaline magmatic intrusive
56
rocks. Additional resources of REY are adsorbed on clay deposits from the weathering and
57
reworking of original primary igneous rocks. China dominates worldwide REY production. The
58
Bayan Obo superlarge deposit currently accounts for approximately 90% of the REE
59
production, with clays accounting for 6–7%.4 In response to the increasing global demand and
60
the mining dominance of China, alternative sources of REY have become a necessity for other
61
countries. Recycling in-use stocks can be an alternative source, especially for the “big four,”
62
i.e., La, Ce, Nd and Pr. The availability of less abundant REE, however, continues to be a
63
challenge.5
64
Noack et al.6 reported a comprehensive study of the distribution of REE in groundwater, lakes,
65
rivers and oceans. According to these authors, REE concentrations in natural waters with a
66
dispersion of one to two orders of magnitude show medians of 5, 53, 71 and 170 pmol/L for
67
the ocean, groundwater, rivers and lakes, respectively. A strong correlation between high REE
4 ACS Paragon Plus Environment
Page 5 of 33
Environmental Science & Technology
68
concentrations and acidity has been reported both in surface and ground waters as well as in
69
leaching studies of soils with different pH solutions.6-9 REE contents reported in acid mine
70
drainage (AMD)9-12 and from our unpublished data consistently show a range between 4,000
71
and 80,000 pmol/L, several orders of magnitude above the medians of natural waters (Figure
72
1).
73
When concentrations are normalized to the North American Shale Composite (NASC),13 the
74
REY distribution patterns in AMD usually show a typical upward convex curvature, indicating
75
enrichment in middle-REE.11,12,14-17 This enrichment is consistent with the middle-REE-depleted
76
pattern shown by the residual oxides resulting from supergene alteration of sulfide
77
deposits.7,18 Light and heavy REY enrichments have also been described and discussed in some
78
AMDs.19,20,21,22
79
Verplanck et al.23 noted that REE behave conservatively in two acid streams at a pH below 5.1
80
and are partitioned into the solid phases at pH values between 5.1 and 6.1. These authors
81
confirmed these observations by neutralizing six AMD samples in the laboratory. This result is
82
also consistent with the decrease in REE observed downstream of the confluence of an AMD
83
with a pristine tributary and the consequent pH increase from 3 to 6 in Lousal mine, Portugal.12
84
In recent decades, passive remediation systems have been implemented to treat acid
85
drainages from coal mines. Among these systems, constructed wetlands with an organic
86
substrate to support bacterially mediated sulfate reduction are the most commonly used.24
87
Due to the high land demand of wetlands, more compact alternative designs, mainly based on
88
the neutralization of acidity (up to pH 6) by means of limestone dissolution, have been
89
developed.25-28 These systems have been extended to high-acidity drainages from massive
90
sulfide mines by increasing the porosity of the calcite substrate.29-31 In some cases, additional
91
alkalinity (up to pH 9) to precipitate divalent metals is added by the dissolution of caustic
5 ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 33
92
magnesia. 32-33 All the passive treatments produce solid products that require disposal in
93
classified repositories and long-term management. Following their behavior in streams,12,23 it is
94
expected that REE will be removed during neutralization and accumulated in the solid
95
precipitates of remediation systems.
96
The objective of the present work was to study the behavior of REY in a multi-stage sequential
97
treatment of AMD: 1) determining how much was removed from the solution at each step; 2)
98
identifying the solid phases accounting for REY retention; and 3) pointing out the selectivity of
99
retention of some REY with respect to the rest.To this end, AMD from a mine gallery (rock
100
leaching) and a roasted pyrite tailing dam (minor rock proportion) were sequentially treated
101
with two successive columns—the first of calcite and the second of caustic magnesia—over 30
102
weeks . Both the pore water and solid phases were analyzed. The results were compared with
103
those for a full-scale treatment, and the economic perspectives of exploitation are discussed.
104 105
EXPERIMENTAL SECTION
106
Water chemistry. Two different AMD samples from the Iberian Pyrite Belt (IPB), SW Spain,
107
were used in the laboratory experiments. The first was sampled at the outflow of a gallery of
108
the Monte Romero mine and represented a strong interaction with a shale enclosing rock. The
109
other was obtained from the leaching from Almagrera mine tailing dam, closed in 2010, and
110
consisted of a material originally dominated by pyrite with minor proportions of silicate
111
minerals.
112
Column setup. Four columns were constructed of transparent polymethyl methacrylate (9.6
113
cm in inner diameter and 40 cm in height) and equipped with lateral ports placed at 5 cm
114
intervals for the water sampling. The columns contained a perforated drain pipe and a 2.5 cm
115
layer of glass beads (3 mm in diameter) at the bottom. Two columns were filled to a height of 6 ACS Paragon Plus Environment
Page 7 of 33
Environmental Science & Technology
116
20 cm with dispersed alkaline substrate (DAS) consisting of calcite sand (0.5-2 mm) and fresh
117
pine wood shavings (5-20 mm) mixed at a 1:1 ratio in weight.34 Calcite dissolution increases pH
118
to 6 and trivalent metals (Al, Fe(III)) precipitate. Wood shavings provide an inert matrix to
119
allocate the neo-formed solids preventing clogging. The other two columns were filled with a
120
DAS consisting of caustic magnesia sand (0.5-1 mm) and fresh pine wood shavings at a 1:1 ratio
121
in weight.34. Caustic magnesia dissolution rises pH to 9, and divalent metals (Zn, Mn, …)
122
precipitate. A decantation vessel was located under each column.
123
During the experiment, input water (i.e., AMD from Monte Romero and Almagrera) was fed by
124
a peristaltic pump to the top of the columns, which were open to the atmosphere. The AMD
125
flowed downward through the substrate and out of the drain pipe into the decantation vessel.
126
The columns were operated at a constant rate of 0.7 mL/min (equivalent to 0.144 m3/m2·day),
127
four times lower than the rates expected in a full-scale treatment.35 A sketch of the system is
128
provided in Figure 2. Water samples were collected from the input container, the supernatant
129
water, and the drain pipe approximately every two weeks. Samples from the intermediate
130
sampling ports along the columns were taken monthly.
131
Chemical and mineralogical analyses. The pH, redox potential and total alkalinity were
132
measured in situ. Filtered samples (0.1 μm nylon) were acidified with HNO3 for analysis of
133
major and trace elements by Inductively Coupled Plasma-Optical Emission Spectroscopy
134
(ICPOES) and Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS), respectively. Details of
135
the analytical procedures are described in Section S1 of the Supporting Information (SI).
136
Once the experiment was concluded, the columns were dismounted, divided in 2 cm-thick
137
slices and dried at room temperature. The solid samples consisted almost completely of
138
residual calcite and organic material (wood chips) with Al and Fe(III)-phases. The mineralogy
7 ACS Paragon Plus Environment
Environmental Science & Technology
Page 8 of 33
139
was determined by X-Ray Diffraction (XRD) and Field Emission Scanning Electron Microscopy
140
with Energy Dispersive Analysis (FESEM-EDS).
141
To enable the study of the partitioning of Al, Fe, and REY into the solid phase in detail, the
142
following sequential extraction procedure adapted from Torres and Auleda36 was applied to
143
the solid residues: 1) water-soluble fraction, extracted with Milli-Q water; 2) basaluminite and
144
calcite, extracted with 1 M ammonium acetate at pH 4.5/20°C; 3) low-crystalline
145
Fe(III)oxyhydroxides, extracted with 0.2 M ammonium oxalate at pH 3/20°C; 4) crystalline
146
Fe(III)oxides, extracted with 0.2 M ammonium oxalate at pH 3/80°; and 5) the residual fraction,
147
digested with a mixture of concentrated HNO3 and HClO4. The detailed protocol can be found
148
in the SI.
149
Synchrotron-based μ-X-ray Fluorescence mapping (µXFM) analyses of selected samples were
150
performed using the ID21 beamline at the European Synchrotron Radiation Facility in Grenoble
151
(ESRF, France). For this experiment, the excitation energy was selected with the use of a Si 111
152
double-crystal monochromator at 7 keV and the beam was focused to a lateral resolution of 1
153
μm with a fixed curvature KB mirrors. The XRF signal was detected using an SGX 100 mm2
154
active area SDD detector. 2D maps were acquired by raster scanning the sample areas with a
155
step size of 1 µm2 and an acquisition time of 100 ms. The XRF spectrum of each pixel of the 2D
156
images was analyzed with PyMCA software, and then elemental maps were obtained through
157
a batch treatment.37
158 159
RESULTS AND DISCUSSION
160
The REY concentration in the Monte Romero AMD is twice that in the Almagrera AMD (Table
161
1). This pattern is consistent with the observation that the concentration of REE in the country
8 ACS Paragon Plus Environment
Page 9 of 33
Environmental Science & Technology
162
rock of the Sitai coal mine (China) is one order of magnitude higher than those of pyrite and
163
coal samples.38
164
As previously described for AMD treatments with the calcite-DAS reactive mixture, three main
165
zones were visually distinguished in the first two columns of both the Almagrera and Monte
166
Romero samples.29,31,35 From the input downstream, these zones are as follows: a red zone
167
characterized by schwertmannite precipitation, a white zone characterized by basaluminite
168
formation, and a gray zone consisting of gypsum and unreacted calcite remaining from the
169
original reagent (Figure S1 of SI). After decantation, the solutions were treated with a MgODAS
170
mixture. A white precipitate of gypsum and Zn(OH)2 was formed close to the water entrance in
171
both treatment lines, with unreacted MgO in the rest of the column (Figure S1).
172
Aqueous phase composition. Depth profiles of the two calcite-DAS columns and the first few
173
centimeters of the MgO-DAS columns after 18 weeks of operation are shown in Figure 3. The
174
numerical values are compiled in Tables S1 and S2 of the SI. The pH values at the outlet of both
175
calcite columns ranged between 5.5 and 6. The pH increase occurred from the beginning of the
176
alkaline interaction in the Monte Romero column, whereas it started to increase after 10 cm in
177
the Almagrera column. This difference is attributed to the exhaustion of the calcite reagent in
178
the second case due to the much higher acidity of the inflow water (Figure S1). Thus, according
179
to the definition of Cravotta9, the trivalent-metal acidities of the inflow waters were 1780 and
180
6200 mg/L CaCO3 for Monte Romero and Almagrera, respectively. In both treatment lines, the
181
pH in the first decantation vessel rose to approximately 6.3 due to CO2 degassing.
182
Downstream, the pH increased to 9.2 in the first 2 cm of both MgO-DAS columns (Figure 3).
183
Finally, due to CO2 dissolution and aragonite precipitation, the pH decreased to 6.8 and 7.5 in
184
the second decantation vessel of the Monte Romero and Almagrera columns, respectively.
9 ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 33
185
The dissolved Fe in the inflow water was primarily Fe(III) in both treatment lines. Most of the
186
dissolved Fe(III) was precipitated in the top centimeters of the Monte Romero column,
187
whereas it required 20 cm to precipitate in the Almagrera column due to the calcite exhaustion
188
(Figure 3).
189
Aluminum was more heavily depleted from the solution in the deeper zones than Fe(III). In
190
Monte Romero, most Al was eliminated in the top 10 cm of the column, whereas 20 cm was
191
required in the Almagrera column (Figure 3). In the Almagrera column, Al initially increased to
192
concentrations higher than in the inflow water. This is a common observation in calcite-DAS
193
treatments32 and can be attributed to basaluminite re-dissolution. Thus, once calcite is
194
exhausted, schwertmannite precipitates at the expense of the alkalinity produced from the
195
dissolution of the previously precipitated basaluminite. Silica followed a pattern similar to Al.
196
Calcium concentrations increased in the depth interval where Al and Fe removal was highest,
197
indicating that calcite dissolution was directly linked to Al- and Fe-hydrolysis and precipitation.
198
As expected, the Mg concentration was constant throughout the calcite-DAS columns and
199
increased to double in the MgO-DAS columns (Tables S1 and S2 of SI).
200
Similarly to previous experiences,29, 32,33 Zn, Mn, Cd, Co and Ni did not change in concentration
201
in the calcite-DAS but were completely removed in the top centimeters of the MgO-DAS
202
columns (Figure 3 for Zn; Tables S1 and S2 for the other metals). The sulfate concentration was
203
partially depleted due to the precipitation of gypsum.
204
REY were removed below detection limits from both treated waters. The first REY decrease
205
coincided with the Al precipitation front. Thus, approximately 35% of the REY in Monte
206
Romero and 55% in Almagrera were captured by the Al-rich solid phase. Then, a second
207
depletion occurred along the transit through the unreacted calcite section of the column: 15%
208
in Monte Romero and 20% in Almagrera. A third drastic removal of REY occurred in the
10 ACS Paragon Plus Environment
Page 11 of 33
Environmental Science & Technology
209
decantation vessel: 40% in Monte Romero and 25% in Almagrera. Finally, the remaining 10%
210
and 5% REY were removed in the top centimeters of the MgO-DAS column in both treatment
211
lines. Among the other trace elements analyzed, only Cu showed a removal pattern very
212
similar to that REY (Figure 3). The small increase in concentration observed for Cu and REY in
213
the pore water from the red zone is attributed to the dissolution of previously precipitated
214
phases.
215
Another interesting issue is the possible fractionation of REE along the different removal steps.
216
Both AMD samples treated here show the characteristic enrichment in middle-REE with
217
respect to the NASC. The sample from Monte Romero, which leaches metasediments, showed
218
a negative Eu anomaly that was less evident in the seeps from the Almagrera tailings. The
219
different REE were consistently depleted from the solution, and no preferential decrease in the
220
concentration of a particular REE was observed along the different stages of treatment (Figure
221
S2 of SI).
222
Solid-phase composition. The distributions of the Fe, Al, ΣREY, Cu and Zn concentrations for
223
each solid sample, obtained via sequential extraction, are shown in Figure S3 of the SI. The
224
complete database, including the different REE and other trace elements, can be found in
225
Tables S3 and S4.
226
The results confirm the major zonation deduced from the pore water. Thus, from surface to
227
bottom, Fe-rich and Al-rich zones with a distinct segregation between them develop in both
228
calcite columns. The zones are closer in the Monte Romero column than in the Almagrera
229
column due to the lower acidity of the former. According to the major mineralogy, Fe is mainly
230
dissolved during steps III and IV of the sequential extraction (reductive dissolution of low
231
crystalline and crystalline oxides). Aluminum is extracted in steps II and III (dissolution at pH
11 ACS Paragon Plus Environment
Environmental Science & Technology
Page 12 of 33
232
4.5 and 3, respectively). The dissolution of a pure basaluminite phase synthesized in the
233
laboratory shows the same pattern (Figure S2 of SI).
234
REY and especially Cu match the Al zone in the Almagrera column. In the Monte Romero
235
column, both Cu and REY are also present in the lower part of the calcite column and in the
236
decantation vessel, where Al is absent. Finally, Zn (as well as Mn, Cd, Ni and Co) is mainly
237
present in the decantation vessel and in the top 2 cm of the MgO columns and is released in
238
steps I and II of the sequential extraction (dissolution of easily soluble phases in water or
239
slightly acidic conditions and carbonates). It should be noted that the concentration in the
240
decantation vessel is higher because the solid is entirely formed by precipitates with no calcite
241
or wood shavings.
242
REY and Cu retention mechanisms. Their high concentrations in pore water and their scarcity
243
in the digestion solutions of the red zone indicate that no significant amounts of Cu and REY
244
were retained in schwertmannite. This finding is consistent with the conservative behavior of
245
Cu and REY in streams at pH below 5,12,23 but it contradicts the REE sequestration by ochre
246
sludge described in wetlands, although no detailed description of the ochre phases is given.39
247
The conservative Cu and REY behavior in the red zone of the treatments leads to a relevant
248
consequence from a practical point of view: REY and Cu are not expected to disperse into
249
schwertmannite precipitating along the stream beds before the AMD is treated. Therefore,
250
both REY and Cu can be entirely recovered in the remediation systems.
251
The first decrease in REY concentration occurred concomitantly with the Al decrease and
252
basaluminite formation. No data on REY sorption in basaluminite are available in the literature.
253
However, the synchrotron-based μ-XRF has enabled the elemental relative abundance to be
254
determined and reveals correlations between the different elements. Thus, a distinct spatial
255
correlation of Al and Ce is evident in solid samples from the basaluminite-rich zone of both
12 ACS Paragon Plus Environment
Page 13 of 33
Environmental Science & Technology
256
columns (Figure 4A, B). This correlation strongly suggests that sorption and co-precipitation of
257
REY in basaluminite occurs.
258
In addition, Cu experienced a decrease with Al in both treatment lines studied. The adsorption
259
of Cu in basaluminite was observed for the AMDs from the Thuringian slate mining area.40 The
260
authors showed that approximately 40% of Cu was retained in Al precipitates when AMD
261
mixed with neutral tributaries. Similar to our column experiments, no other divalent metal was
262
retained in this mixing process.
263
The REY and Cu concentrations in the pore water continued to decrease after Al was practically
264
absent, suggesting that a mechanism of retention other than basaluminite precipitation was
265
occurring. Indeed, DRX and FESEM-EDS observations showed the precipitation of gypsum
266
inside the calcite column, the decantation vessel and the top 2 cm of the MgO columns.
267
Together
268
(Zn,Cu)Zn(OH)13[(Si,S)(O,OH)4] and small cubes of cuprite (Cu2O) were observed and identified
269
by XRD analysis in samples downstream of the Al zone. These minerals can account for the Cu
270
depletion in the pore water.
271
In contrast to the basaluminite samples, µ-XRF mapping of samples from the end of the calcite
272
columns and decantation vessels showed no correlation between REE and Al, which seemed to
273
be enriched in distinct discrete particles. Indeed, spherical aggregates and plates of REE-Ca-
274
Fbearing phase (fluorite?) were also observed together with bechererite in samples from the
275
end of the calcite column, the decantation vessel and the top 2 cm of the MgO column of
276
Monte Romero (Figure 4D, E). Fluorite was also identified by DRX (Figure S4 of SI).
277
Fluorite has been investigated as a host for REE in magmatic and hydrothermal deposits. Thus,
278
fluorite veins have been selected as a possible target for REE exploration in Australia,41 China,42
279
the USA44 and South Africa.43 However, to the authors’ knowledge, no REE-fluorite formation
with
gypsum,
aggregates
of
trigonal
pyramids
of
bechererite
13 ACS Paragon Plus Environment
Environmental Science & Technology
Page 14 of 33
280
has been described at environmental temperatures. The precipitation of fluorite in the Monte
281
Romero treatment has two causes. Firstly, calcite dissolution supplies the Ca2+ required for
282
fluorite formation. However, no fluorite is found in the schwertmannite zone, where active
283
calcite dissolution occurs. Secondly, in AMD, F- is entirely complexed with Al3+ as AlF2+ (Figure
284
S5 of SI). When Al is precipitated as basaluminite, the complex becomes unstable and F is
285
mainly found as the free anion F-; the saturation index of fluorite is reached, and the mineral
286
precipitates. This process may occur in some pores of the Monte Romero column but is more
287
acute in the decantation vessel. The Almagrera AMD is much richer in Mg, and a strong MgF+
288
complex forms together with free F- when Al is depleted (Figure S5) and fluorite is scarcer.
289
The pore water geochemistry was, in general, consistent with the observed mineralogy. Thus,
290
the pore water from the deeper part of the calcite column of both treatment lines was
291
supersaturated in basaluminite and cuprite, in equilibrium with gypsum and subsaturated in
292
calcite (Figure 5). In partial disagreement with the observations, however, fluorite saturation
293
was only reached at the first decantation vessel of the Monte Romero line. This discrepancy
294
may have been due to the effect of REY on the fluorite solubility product, although no data on
295
this point have been found in the literature. Although supersaturation was exhibited by the
296
pore waters, no fluorapatite was observed under FESEM-EDS inspection, and moreover, no P
297
was detected in any step of the sequential extraction of the solid phase. Therefore, although it
298
cannot be completely ruled out, phosphates do not seem to play a significant role in REY
299
retention.
300 301
ECONOMIC IMPLICATIONS
302
As shown above, no REY were partially lost in schwertmannite precipitates during the AMD
303
transit through streams or pre-treatment lagoons, and all dissolved REY reached the 14 ACS Paragon Plus Environment
Page 15 of 33
Environmental Science & Technology
304
remediation system. Indeed, practically all REY were completely retained in the basaluminite
305
layer and the decantation vessels of the calcite-DAS passive remediation systems. Therefore,
306
the REY content can be entirely recovered from AMD. Several full-scale treatment systems are
307
under construction in the IPB. However, only one full-scale system at Mina Esperanza and a
308
pilot plant at Monte Romero were operating for more than one year until the calcite reagent
309
was exhausted.32,34 Both systems reproduced the same zonation pattern described for the
310
column experiments: schwertmannite, basaluminite and calcite-gypsum zones (Figure S5 of SI).
311
As described for column experiments, the segregation between zones was also distinct in the
312
two field cases. 32,34 The annual reserves and rate can be roughly approximated from the total
313
dissolved Al and REY concentrations in the AMD and the molar volume of basaluminite. An
314
example for the two sites studied is shown in Table 2. The rates calculated from the AMD
315
concentrations are similar to the analysis performed in the basaluminite residues excluding
316
wood shavings: 0.30% for Mina Esperanza and 1.44% for Monte Romero. It is interesting to
317
note that U and Th concentrations were very low (