Subscriber access provided by TULANE UNIVERSITY
Sustainability Engineering and Green Chemistry
Comprehensive understandings of rare earth element (REE) speciation in coal fly ashes and implication for REE extractability Pan Liu, Rixiang Huang, and Yuanzhi Tang Environ. Sci. Technol., Just Accepted Manuscript • Publication Date (Web): 26 Mar 2019 Downloaded from http://pubs.acs.org on March 26, 2019
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 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 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.
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 29
Environmental Science & Technology
1
Comprehensive understandings of rare earth element (REE) speciation in coal fly ashes
2
and implication for REE extractability
3 4
Pan Liu, Rixiang Huang, Yuanzhi Tang*
5 6
School of Earth and Atmospheric Sciences, Georgia Institute of Technology, 311 Ferst Dr,
7
Atlanta, GA 30332-0340, USA
8 9
* Corresponding author.
10
Email:
[email protected]; Phone: 404-894-3814
11 12 13 14
Submit to
15
Environmental Science & Technology
1 ACS Paragon Plus Environment
Environmental Science & Technology
16
Page 2 of 29
Abstract
17
In recent years, recovery of rare earth elements (REE) from coal fly ashes (CFAs) has
18
been considered as a promising resource recovery option. Yet, quantitative information on REE
19
speciation in CFAs and its correlation with REE extractability are not well established. This
20
study systematically investigated REE speciation-extractability relationship in four
21
representative CFA samples by employing multiple analytical and spectroscopic techniques
22
across micro to bulk scale and in combination with thermodynamic calculation. A range of
23
REE-bearing phases are identified, including REE oxides, REE phosphates, apatite, zircon, and
24
REE-bearing glass phase. REE can occur as discrete particles, as particles encapsulated in glass
25
phase, or distribute throughout the glass phase. Although certain discrepancies exist on the
26
REE speciation quantified by X-ray adsorption spectroscopy and acid leaching due to intrinsic
27
limitations of each method, both approaches show significant fractions of REE oxides, REE
28
phosphates, apatite, and REE-bearing Fe oxides. This study contributes to an in-depth
29
understanding of REE speciation-distribution-extractability relationship in CFAs and can help
30
identify uncertainties associated with the quantification of REE speciation. It also provides a
31
general methodology for future studies on REE speciation in complex environmental samples
32
and a knowledge basis for the development of effective REE recovery techniques.
2 ACS Paragon Plus Environment
Page 3 of 29
33
Environmental Science & Technology
1. Introduction
34
Rare earth elements (including lanthanides and yttrium, as REE hereafter) are critical
35
for a wide range of high tech applications due to their unique physicochemical properties.1 The
36
growing importance of REE to technologies and economies and the potential risks of supply
37
disruptions have led both the United States (US) and the European Union to label REE as
38
“critical materials”.2,
39
importing REE in 2017.4 Over the recent years, there are great interests in developing
40
economically feasible and environmentally friendly approaches for domestic REE recovery.5-7
41
With this regard, coal combustion products such as coal fly ashes (CFAs) have recently
42
emerged as a promising REE source.8-11 In the US, around 45 Mt CFAs are generated
43
annually,10 with only 39% beneficially used and the rest discarded.12 On the other hand, REE
44
are enriched in the CFAs after coal burning.8, 13 With an average total concentration of REE in
45
CFAs at ~500 ppm, the annual value of REE derived from CFAs is estimated to be $4.3 billion
46
in the US.10 Thus, recovery of REE from CFAs is a promising waste recycling option that might
47
bring about significant environmental, economic, and societal benefits.
3
The US is currently a net importer of REE and spent $150 million
48
In order to develop cost effective and environmental friendly REE recovery techniques,
49
many critical issues need to be addressed,11 including the following two fundamental questions:
50
(1) the speciation of REE in CFAs, such as the chemical species and physical distribution
51
and/or association with other phases, and (2) the controlling factors on REE extractability.
52
Several previous studies14-16 have suggested that REE are dispersed throughout the amorphous
53
aluminosilicate glass phase, thus the glass fraction should be targeted for REE recovery from
54
CFAs. Other studies suggested the (predominant) presence of REE in phosphate (e.g.,
55
monazite)17, carbonate18, and/or silicate minerals (e.g., zircon)15. Such inconsistency might
56
result from the intrinsic heterogeneity of CFAs as well as the limitations of individual analytical
57
method. 3 ACS Paragon Plus Environment
Environmental Science & Technology
Page 4 of 29
58
Both micro- and bulk-scale analytical methods have been utilized to study the
59
correlation between REE speciation (e.g. chemical species and physical distribution) and
60
extractability in environmental samples, including CFAs. The main techniques used to study
61
REE speciation include scanning/transmission electron microscopy coupled with energy
62
dispersive X-ray spectroscopy15, 17-19, electron probe microanalysis19, laser ablation inductively
63
coupled plasma mass spectrometry15, 20, and synchrotron X-ray microscopy and spectroscopy21,
64
22.
65
and acid leaching26. Although these techniques can provide valuable information (e.g., REE
66
species, distribution, and extractability) at different spatial/temporal scale, they also possess
67
intrinsic limitations. For example, direct information on REE extractability can be provided by
68
sequential chemical extraction, but interpretation of the related mineral phases might be
69
problematic, because this empirical method might cause incomplete dissolution of desired
70
phases, concomitant dissolution of different phases, or alteration of the original phases. To
71
obtain a more complete picture and quantitative information of REE speciation in complex
72
matrices such as CFAs, comparison and cross-validation of information obtained by different
73
techniques are highly desired.
Extractability evaluations were mainly conducted using sequential chemical extraction23-25
74
The goals of this study are to comprehensively characterize and quantify REE species
75
and distribution in CFAs using complementary techniques across different spatial scales, and
76
to correlate the speciation information with REE extractability. Four representative CFA
77
samples were selected by taking into consideration of both CFA chemical composition and
78
main coal basins in the US. A systematic method was developed by employing techniques
79
across bulk to molecular scales, including chemical (acid leaching and sequential extraction),
80
mineralogical (X-ray diffraction), microscopic (scanning electron microscopy coupled with
81
energy dispersive X-ray spectroscopy), and spectroscopic methods (synchrotron X-ray
82
adsorption spectroscopy). Based on the results obtained, the advantages and disadvantages of 4 ACS Paragon Plus Environment
Page 5 of 29
Environmental Science & Technology
83
each technique and controlling factors on REE extractability were evaluated. The results
84
contribute to an improved understanding of REE occurrence in CFAs and their extractability,
85
evaluate a general methodology for characterizing REE speciation in complex environmental
86
matrix, and provide a knowledge basis for the development of cost effective and
87
environmentally friendly REE recovery methods.
88 89
2. Materials and methods
90
2.1 Materials
91
The American Society for Testing and Materials (ASTM) C 618 defines two classes of
92
fly ashes based on chemical characterization: Class F CFAs (with SiO2 + Al2O3 + Fe2O3 ≥ 70
93
wt%) and Class C CFAs (with 50 wt% ≤ SiO2 + Al2O3 + Fe2O3 ≤ 70 wt%).27 Based on this
94
classification and the major coal basins in the US, four representative CFA samples were
95
selected for this study (Table 1), including two CFA samples (F-1 and C-1) from a local coal
96
fired power plant located in the southeastern US28 and two CFA samples from the National
97
Institute of Standards and Technology (NIST) (SRM-2690 as F-2 and SRM-2691 as C-2
98
hereinafter). Samples F-1 and F-2 are Class F CFAs and their feed coals are bituminous coal
99
from the Illinois Basin and subbituminous coal from a surface mine in Craig, CO, respectively.
100
Samples C-1 and C-2 are Class C CFAs and their feed coals are subbituminous coals from the
101
Powder River Basin (details in the Supporting Information SI, Table S1). A suite of REE
102
reference compounds (REE-organic complexes, REE oxides, REE carbonates, REE phosphates,
103
and natural REE-bearing minerals, etc) were obtained or synthesized using chemicals of ACS
104
grade or higher (details in Table S2). The synthesized and natural REE reference compounds
105
were confirmed by X-ray diffraction to be pure phases.
106 107
2.2. Composition analysis and chemical extraction of REE in CFAs 5 ACS Paragon Plus Environment
Environmental Science & Technology
Page 6 of 29
108
REE contents in CFAs were determined from total digestion29. Two types of chemical
109
extraction methods were used to assess the extractability of REE, including sequential
110
extraction and acid leaching. Sequential extraction was conducted following a previous
111
procedure.30 Briefly, four REE fractions are defined: (1) acid-soluble (Step-1, e.g., carbonates),
112
(2) reducible (Step-2, e.g., Fe-Mn oxides), (3) oxidizable (Step-3, e.g., organic matter and
113
sulfide), and (4) residue (e.g., silicates). For acid leaching, REE extractability was investigated
114
as a function of pH adjusted using HNO3 and NaOH. At desired pH, 0.2 g CFAs per 100-mL
115
H2O were reacted for three days under agitations at room temperature. At the end of each
116
chemical extraction step or acid leaching, REE concentration was analyzed using inductively
117
coupled plasma mass spectrometry (ICP-MS). Details on total digestion, sequential extraction,
118
and ICP-MS measurement are in SI Text S1.
119 120
2.3 Mineralogical, microscopic, and spectroscopic analysis of CFAs
121
X-ray diffraction (XRD), scanning electron microscopy with energy dispersive X-ray
122
spectroscopy (SEM-EDX), micro X-ray fluorescence (μXRF) imaging and micro X-ray
123
absorption near edge structure spectroscopy (μXANES), and bulk XANES analyses were
124
conducted on the CFA samples. For XANES analyses, Nd and Y were chosen as representative
125
light REE (LREE, from La to Sm) and heavy REE (HREE, from Eu to Lu plus Y), respectively.
126
μXRF imaging and μXANES spectra of Y K-edge, Nd LIII-edge, and Nd LII-edge were
127
collected at Beamline 2-3 of the Stanford Synchrotron Radiation Lightsource (SSRL, Menlo
128
Park, CA). Bulk XANES spectra were collected at Beamlines 5-BM-D at the Advanced Photon
129
Source (APS, Lemont, IL). Spectra of reference compounds were also collected for liner
130
combination fitting (LCF) analysis. Details on XRD, SEM-EDX, μXRF, and bulk/micro
131
XANES data collection and LCF analysis are in SI Text S2.
132 6 ACS Paragon Plus Environment
Page 7 of 29
Environmental Science & Technology
133
3. Results and discussion
134
3.1 Chemical and mineralogical characteristics of CFAs
135
The element contents of CFA samples are listed in Tables 1 and S1 and compared to
136
common CFAs from the major coal basins in the US. Samples F-1 and F-2 show chemical
137
composition typical of Class F CFAs that are enriched in SiO2 (55%), Al2O3 (25%), and Fe2O3
138
(5–12%), while samples C-1 and C-2 as Class C CFAs are relatively depleted in the above
139
elements but abundant in alkaline oxides (26–28% CaO and 1–5% MgO). Total REE content
140
in Class F CFAs is similar to that of Class C CFAs (250–320 ppm), and comparable to common
141
CFAs in the US (Table S1). REE patterns of the four CFA samples normalized (with subscript
142
N) by the Upper Continental Crust (UCC, representing the natural REE background)31 are
143
shown in Fig. S1, which are in line with the general REE patterns of CFAs. REE are M-type
144
enriched (LaN/SmN < 1 and GdN/LuN > 1)8 in both Class F and C CFAs without obvious Ce
145
anomaly (defined as 𝐶𝑒 ∗ = 2𝐶𝑒𝑁 (𝐿𝑎𝑁 + 𝑃𝑟𝑁)). Samples C-1, C-2, and F-2 are characterized
146
with positive Eu anomaly (𝐸𝑢 ∗ = 2𝐸𝑢𝑁 (𝑆𝑚𝑁 + 𝐺𝑑𝑁)), while sample F-1 does not show
147
evident anomaly.
148
The mineralogical composition of CFAs was investigated by XRD (Fig. S2).
149
Crystalline phases identified in Class F CFAs include quartz (SiO2), mullite (Al6Si2O13),
150
hematite (Fe2O3), and magnetite (Fe3O4). Class C CFAs have a more complex mineralogy, with
151
quartz, anhydrite (CaSO4), tricalcium aluminate (Ca3Al2O6), lime (CaO), and periclase (MgO)
152
being identified. A broad hump at around 20–30 2 suggests the presence of amorphous
153
aluminosilicate glass, a common major component (50–80 wt%) in CFAs.32 SEM analysis
154
observed different morphologies of CFA particles, with the predominance presence of
155
spherical and cenospherical particles ranging at 1–100 μm (Fig. S3). Such morphological
156
features are typical for CFAs due to particle melting and decomposition during coal burning
157
process.33 In addition, considerable amounts of fine particles (e.g., Fe oxides) were found to be 7 ACS Paragon Plus Environment
Environmental Science & Technology
158
Page 8 of 29
embedded within glass phase in both Class F and C CFAs (Figs. S3 and S4).
159 160
3.2 REE speciation by SEM-EDX
161
SEM-EDX observed two types of REE occurrence in CFAs, including pure REE
162
phosphates and minor REE in lime and zircon. Fig. 1a shows a REE phosphate particle (~30-
163
μm size) that is enriched in LREE, while the one in Fig. 1b (~10-μm size) is enriched in HREE
164
with close association with aluminosilicate glass. These two REE phosphate particles might be
165
monazite/rhabdophane (LREE-enriched phosphate) and xenotime/churchite (HREE-enriched
166
phosphate), respectively. Monazite and xenotime are anhydrous phosphate minerals, while
167
rhabdophane and churchite are hydrated phosphate minerals. It is difficult to differentiate them
168
by SEM-EDX. REE phosphates have also been identified previously using transmission
169
electron microscopy17, 34 and electron probe microanalysis19. A lime (CaO) particle containing
170
minor REE (REE oxide content < 0.5 wt%) is shown in Fig. 1c. This phase contains minor
171
HREE and its spherical shape suggests that it might be relics of REE-bearing CaCO3 after
172
burning. Both REE-bearing CaCO3 and CaO have been previously identified.18, 19 A zircon
173
(ZrSiO4) particle (REE oxide content < 0.3 wt%) is shown in Fig. 1d, which is encapsulated in
174
glass phase. Although zircon is relatively abundant in the CFAs, only a portion of observed
175
zircon particles (4 out of 16) contains EDX-detectable REE. Other potential REE-bearing
176
phases, such as apatite19 and glass phase14, were observed but with no REE detected by EDX.
177
An overview of the REE speciation observed by SEM-EDX reveals three critical
178
features regarding REE occurrence in CFAs (Table S3). First, different phases might
179
preferentially
180
monazite/rhabdophane, and HREE in xenotime/churchite, lime, and zircon (Fig. 1). This
181
phenomenon might be due to the gradually decreasing cation radii of REE with increasing
182
atomic number (i.e., lanthanum contraction), which changes the compatibility of individual
accommodate
LREE
or
HREE.
LREE
are
more
enriched
in
8 ACS Paragon Plus Environment
Page 9 of 29
Environmental Science & Technology
183
REE element for a specific mineral phase. Second, as shown in Figs. 1 and S5, REE-bearing
184
phases can occur as discrete particles or be closely associated with glass phase (at the edge or
185
totally encapsulated). The close association of REE-bearing phases with glass phase is likely
186
due to the capture of REE-bearing phases during glass cooling and agglomeration21, and it has
187
also been noticed by several previous studies (e.g., ref.17, 21). Third, REE-bearing phases might
188
be subjected to decomposition during combustion. Although REE phosphates and zircon are
189
typically thought to be stable even at high temperature (e.g., ref19), Fig. S6 shows
190
decomposition characteristics of REE phosphates (particle shattering) and zircon (holes and
191
melting features), as compared to Figs. 1a and 1d, respectively.
192 193
3.3 REE speciation by synchrotron microscopy and spectroscopy
194
Characteristics of REE reference compounds. Y XANES spectra of reference
195
compounds are plotted in Figs. S7. Spectra of Y2O3, Y-churchite (YPO4·H2O), apatite
196
(Ca5(PO4)3(OH, F)), REE-bearing calcite (CaCO3), monazite ((Ce, Th, La, Nd)PO4), xenotime
197
((Y, Dy)PO4), zircon (ZrSiO4), and REE-bearing glass all show a peak at 17056.5 eV and a
198
shoulder at 17066 eV, but the relative intensities are different, and with additional features at
199
17080–17140 eV. These different spectral features reflect different structures and local
200
coordination environments of the REE reference compounds (such as the first shell
201
coordination, summarized in Table S2). On the other hand, Y spectra of Y3+(aq), Y-tengerite
202
(Y2(CO3)3·2-3H2O), bastnäsite ((Ce, La, Nd)CO3(OH, F)), REE-bearing hematite (Fe2O3), and
203
REE-organic complexes are similar to each other, showing a single peak at around 17056.5 eV,
204
although the peak position is slightly right-shifted for Y3+(aq) and left-shifted for bastnäsite.
205
Nd spectra were not collected for some reference compounds (e.g., bastnäsite and
206
monazite) due to interference of Ce LII-edge signal on Nd LIII-edge and Sm LIII-edge signal on
207
Nd LII-edge. For reference compounds with Nd spectra obtained, their LIII- and LII-edge spectra 9 ACS Paragon Plus Environment
Environmental Science & Technology
Page 10 of 29
208
are generally similar in appearance with a characteristic intense main peak (Fig. S8), making it
209
difficult to distinguish REE speciation based on Nd XANES. Thus, Y was used as the primary
210
“probe” for characterizing REE speciation. Although some Y spectra of REE-reference
211
compounds are similar, giving the possibility that some subtle differences might still be helpful
212
in differentiating REE speciation, all thirteen Y reference spectra were used for the linear
213
combination fitting (LCF).
214
μXRF and μXANES analyses. Representative µXRF maps were plotted in Fig. 2 and
215
SI Figs. S9 and S10. Most Y hotspots observed have a particle size of 60%)
229
at those hotspots by LCF. Five hotspots are identified as REE oxides (Table S4) and they have
230
a relatively smaller particle size (generally 3.5 (REE oxides, carbonates)
pH 3.51.5 (apatite)
pH