Environmental Pollution 244 (2019) 174e181

Contents lists available at ScienceDirect

Environmental Pollution
journal homepage: www.elsevier.com/locate/envpol

Mechanism of uranium release from uranium mill tailings under long-

term exposure to simulated acid rain: Geochemical evidence and
environmental implication*
Meiling Yin a, 1, Jing Sun b, 1, Yongheng Chen a, Jin Wang a, c, d, *, Jianying Shang e,
Nick Belshaw d, ChuanChou Shen f, g, Juan Liu a, d, **, Huosheng Li a, Wensheng Linghu h,
Tangfu Xiao a, Xinjiao Dong i, Gang Song a, c, Enzong Xiao a, Diyun Chen a, c
a
School of Environmental Science and Engineering, Guangzhou University, Guangzhou, 510006, China
b
School of Earth Sciences, University of Western Australia, Perth, WA, 6009, Australia
c
Guangdong Provincial Key Laboratory of Radionuclides Pollution Control and Resources, Guangzhou, 510006, China
d
Department of Earth Sciences, University of Oxford, Oxford, OX1 3AN, UK
e
College of Resources and Environmental Sciences, China Agricultural University, Beijing, 100193, China
f
Department of Geosciences, National Taiwan University, Taipei, 10617, Taiwan
g
Research Center for Future Earth, National Taiwan University, Taipei, 10617, Taiwan
h
College of Chemistry and Chemical Engineering, Shaoxing University, Shaoxing, 312000, China
i
School of Life & Environmental Science, Wenzhou University, Wenzhou, 325027, China

a r t i c l e i n f o a b s t r a c t

Article history: To date, there is not sufficient knowledge to fully understand the occurrence, transport and fate of re-
Received 24 June 2018 sidual uranium (U) from uranium mill tailings (UMT). Herein this study investigated different U release
Received in revised form behaviors from natural UMT (without grinding) under four simulated acid rain (pH ¼ 2.0e5.0) compared
24 September 2018
with controlled scenario (pH ¼ 6.0) for 25 weeks. The results showed that the most notable U release was
Accepted 3 October 2018
observed from UMTpH2.0, followed by UMTpH3.0 whereas a nonlinear relationship between pH and U
Available online 9 October 2018
release was observed from UMTpH4.0e6.0. The divergence of U release behaviors was attributed to the
presence of minerals such as calcite and clinochlore. Autunite, a secondary mineral formed after leaching,
Keywords:
Uranium
might regulate U release in UMTpH3.0e6.0. Fick theory model revealed the shift of U release mechanism
Uranium mill tailings from surface dissolution to diffusion transport for UMTpH2.0, UMTpH3.0 and UMTpH5.0 at varied stage,
Leaching test whereas UMTpH4.0 and UMTpH6.0 displayed univocal dissolution and diffusion mechanism, respectively.
Acid rain This study highlights the necessity of performing long-term leaching tests to detect the “shift event” of
leaching kinetics and to better understand the mechanism of U release influenced by mineralogy of the
natural UMT under simulated acid rain conditions, which is conducive to developing UMT management
strategies to minimize the risk of U release and exposure.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction

Uranium (U) is a toxic and radioactive contaminant in the

environment; there has been increasing number of studies focusing
*
This paper has been recommended for acceptance by Dr. J Rinklebe.
on the remediation of U contamination (Carter et al., 2009; Pan
* Corresponding author. School of Environmental Science and Engineering, et al., 2016; Rufyikiri et al., 2006) and a variety of new methods
Guangzhou University, Guangzhou 510006, China. of U removal have been developed in accordance (Gu et al., 2018;
** Corresponding author. Innovation Center and Key Laboratory of Waters Safety Han et al., 2018; Yin et al., 2017; Zhang et al., 2018). At the same
& Protection in the Pearl River Delta, Guangzhou University, Guangzhou 510006, time, U has become one of the world's most important energy
China.
E-mail addresses: wangjin@gzhu.edu.cn (J. Wang), juan.liu@earth.ox.ac.uk,
materials in the last 60 years, accompanied with increasing U
liujuan@gzhu.edu.cn (J. Liu). mining and processing activities (Fang et al., 2018; Ilankoon et al.,
1
These authors contributed equally to this publication.

https://doi.org/10.1016/j.envpol.2018.10.018
0269-7491/© 2018 Elsevier Ltd. All rights reserved.
 M. Yin et al. / Environmental Pollution 244 (2019) 174e181 175

2018). Uranium mill tailings (UMT) are the crushed ore residues known deposit of U in China (Luo et al., 2017). Previous studies
from the extraction of U from ores such as pitchblende (Landa, have revealed that the granite hosted U ore deposits provide ~30%
2004). Uranium mining and processing have been concomitant of the U production in China during the past three decades (Hu
with the generation of huge amount of UMT. According to a pre- et al., 2008; Hu et al., 2009; Luo et al., 2015; Wang et al., 2017b).
vious estimate from Steinhausler and Zaitseva (2007), the total The determined U concentration in the fine ore was
amount of UMT due to U mining worldwide is about 2352 million 10957 ± 13 mg/kg, and the reported grade (% U) of the granite
tones. These UMT often contain surprisingly high concentrations of hosted U ore deposit in China is 0.1e1 (Dahlkamp, 2009), which
residual U, thorium and other toxic metals (Robinson, 2004; Zhang shares the same order of magnitude. The natural UMT (without
et al., 2015), and usually have a much higher porosity and perme- grinding) used in this study were collected from an un-
ability than the ores from which they are derived (Ferry et al., rehabilitated tailings dam where huge amount of UMT were
2002). In the past decades, it is not rare that UMT has been piled near a large U hydrometallurgy plant in Guangdong Prov-
stacked on the land surface after simple treatment of lime ince, South China. The freshly disposed UMT with slightly wet
neutralization. Some of the UMT are further treated with landfill surface were sampled using a wooden shovel, stored in a large
cover with a variety of plants, vegetation or trees to enhance the plastic box, and were then air dried prior to semi-dynamic
conservation (Waugh et al., 2008). leaching tests. The grain size of the UMT selected for this study
However, the chemical and radioactive elements in UMT could were between 6 and 10 mm, which is the dominant size fraction of
be mobilized during long-term redox and weathering processes. the UMT. The U concentration in this fraction of the UMT was
The mobilization would be even more significant under acid con- 997.9 ± 52.6 mg/kg. The concentrations of some other closely
dition as acidity is one of the most important factors influencing the related elements were shown as follows: P (0.07%), K (3.3%), Mg
release of toxic elements from solid wastes (Fo €rstner, 1987). The (0.6%), Na (1.7%), Fe (2.0%), Al (6.9%) and Ca (0.7%).
acid condition could be achieved in UMT if sulfuric acid was used to
dissolve the U from the ores. Except for ores having a high car- 2.2. Semi-dynamic leaching test
bonate content, the commercial extraction of U from ores generally
involves leaching with sulfuric acid (Landa, 2003). For most of U To investigate the long-term fate of U in UMT during the inter-
industries in China, sulfuric acid was also mainly adopted (Wang mittent leaching and weathering process, five sets of semi-dynamic
et al., 2012a). The acid condition could also be achieved due to leaching tests with simulated acid rain and normal rain were car-
oxidation of sulfidic minerals in the tailings; the extent of acid ried out. Before the leaching test, UMT was washed with deionized
generation is dependent on the concentration of sulfide, which water three times to remove the surface-adhered salts. The five
ranges from a few to tens of wt% in UMT (Abdelouas, 2006). The low leaching tests with pH ranging from 2.0 to 6.0 with a total duration
pH condition could accelerate the release of radioactive and haz- of 25 weeks, were then conducted as follows. Leaching experiments
ardous elements from UMT. Consequently, the residual U and other were carried out in five brinell funnels (20 cm in diameter) each
metals in the UMT can release into ambient environment, and thus loaded with 1 kg of UMT. In each leaching cycle, the five UMT
UMT may become a substantial source of surface and groundwater loaded brinell funnels were sprayed slowly with a plastic watering
contamination (Liu et al., 2017; Liu et al., 2015; Wang et al., 2016). pot, which resembled the rainfall precipitation process. 800 mL of
The release would be greatly enhanced if suffered from the local the simulated rain was used in each brinell funnel, which could just
acid rain. Many studies have demonstrated the significant effect of inundate the UMT's surface. Deionized water acidified by sulfuric
pH on U release from various materials (Mason et al., 1997; Xu et al., acid (trace metal grade) to pH 2.0, 3.0, 4.0 and 5.0 were used as
2004; Zhang and Wang, 2017). Nevertheless, few studies have simulated acid rain, whereas deionized water (pH ~6.0) was used as
examined the behaviors and mechanisms of U mobilization from simulated normal rain for comparison. The deionized water was
natural UMT (without grinding). Although general leaching char- produced by Milli-Q System (Millipore, Bedford, USA), and no
acteristics can be identified in leaching studies using U tailings additional treatment was used. After soaking for 1 h, the leachate
ground before leaching, identifying and quantifying the natural was recovered for physico-chemical analysis, and the UMT in the
processes of exposure of U-containing minerals to long-term brinell funnels was exposed to the air for the rest of the week until
weathering and corrosion remains a substantial challenge. To next leaching cycle commenced. The leachants were renewed in
date, the knowledge about natural UMT hazards in the light of long- each cycle. The semi-dynamic leaching events were carried out
term leaching studies is limited, which however is essential for the once every 7 days and lasted for 25 weeks. The experiments were
development of sound UMT management strategies. performed in open system to ensure good simulations of the nat-
Semi-dynamic leaching tests are usually employed to elucidate ural weathering conditions for the UMT. The five sets of semi-
the dominant leaching mechanism from a solid waste in which the dynamic leaching tests were performed simultaneously, and were
leaching agent is renewed periodically (Kosson, 2002). Using this referred to hereafter as UMTpH2.0, UMTpH3.0, UMTpH4.0, UMTpH5.0
method, the present study investigated the release of U from the and UMTpH6.0, respectively.
most common natural UMT. The leaching experiments were per- After each leaching cycle, the leachate was drained into
formed under simulated acid rain with pH ranging from 2.0 to 5.0 1000 mL beaker. After settlement for 2 min, the supernatants
and a simulated normal rain for comparison. The objectives of this were filtered to 0.45 mm using nylon syringe filters. The pH, redox
study were (i) to quantify the long-term U leaching characteristics potential (ORP) and conductivity (EC) of the filtered leachate were
from natural UMT influenced by acid rain, (ii) to unravel the measured immediately after collection. The rest of the filtered
mobilization mechanism based on mineralogy and kinetics evi- leachate was acidified to 1% HNO3 (trace metal grade), and stored
dences, and (iii) to inform the prospects for remediating U at 4  C until analysis for solution composition. UMT before and
contamination from UMT. after 25 week leaching were collected and used for mineralogical
and morphological characterizations by X-ray diffraction (XRD),
2. Materials and methods scanning electron microscope and energy dispersive spectrometer
(SEM-EDS), transmission electron microscope and energy
2.1. UMT sampling dispersive X-ray spectrometry (EDS) in scanning TEM (TEM/
STEM-EDS), Fourier-transform infrared spectroscopy (FTIR) and
The U metallogenic province in South China is the largest sequential extraction.
176 M. Yin et al. / Environmental Pollution 244 (2019) 174e181

2.3. Analytical methods for the leachate 2.8) extraction targeting exchangeable and acid-extractable U, (ii) a
16 h hydroxylamine-hydrochloride (NH2OH$HCl, 0.5 M, pH 1.5)
pH and redox potential of the leachate were measured simul- extraction targeting reducible U, and (iii) a 16 h hydrogen peroxide/
taneously using a calibrated redox potential-acidity meter (PB-10, ammonium oxalate (8.8 M H2O2/1.0 M NH3OAc, pH 2.0) extraction
Sartorius, Inc., Germany), and EC was measured using a calibrated targeting oxidizable U. Extractions were conducted with a dry mass
conductivity meter (ECTestr11þ, Eutech Instruments, Inc., USA). of 0.5 g and an extractant volume of 20 mL (25 mL for NH3OAc), at
Dissolved U concentration in the leachate was determined by room temperature in constantly agitated polyethylene centrifuge
inductively coupled plasma mass spectrometry (ICP-MS, Elan 6100 tubes. At the end of each extraction step, suspensions were
DRC-e, PerkinElmer, Inc., USA). Each sample was run three times centrifuged. The supernatants were then filtered to 0.2 mm and
and averaged. Accuracy was verified against known quality con- analyzed by ICP-MS as described above, and solids were treated
trols. The standard deviations of three runs were less than 5%. with the next extractant. The residual fraction after extractions was
Dissolved concentrations of other elements were determined by digested with HNO3-HF-HClO4 (trace metal grade) and analyzed
inductively coupled plasma optical emission spectroscopy (ICP- accordingly.
OES, Optima 7000 DV, PerkinElmer, Inc., Waltham, USA).
2.4.6. Fick theory model
2.4. Analytical methods for UMT The mechanism of U release can be depicted by the slope of the
linear regression of the logarithmic of cumulative flux against
2.4.1. XRD leaching time (Torras et al., 2011). When the slope is > 0.65, the
XRD was used to assess UMT bulk mineralogy. To conduct XRD release of U is dominated by mineral dissolution; when the slope is
analysis, the air dried UMT samples were ground to powder with an between 0.35 and 0.65 (including 0.65), the release of U is domi-
average particle diameter smaller than 5 mm using an agate mortar- nated by diffusion transport; and when the slope is  0.35, the
and-pestle set, and then mounted on a titanium aluminide-alloy release of U is dominated by surface wash-off. The cumulative flux
test block. XRD patterns were collected using a PANalytical X'Pert (Bac i) was calculated based on previously published procedures
Pro diffractometer (The Netherlands). The diffractometer used Cu (Wang et al., 2012b).
Ka radiation (l ¼ 1.5406 Å) and scanned over 2q range from 5 to
80 , with a step size of 0.02 and a counting time of 150 s per step. 3. Results and discussion
The identification search match of the various minerals were ob-
tained using MDI Jade 5.0 equipped with the JCPDS PDF-2 database. 3.1. The nonlinear effects of leaching pH

2.4.2. SEM-EDS Dissolved concentration of U and leachate pH from the semi-

SEM-EDS was used to assess the morphology and composition dynamic leaching tests are presented in Fig. 1, where four simu-
of the pre- and post-leaching UMT powder. To avoid the charging lated acid rain and one simulated normal rain were compared.
effect on the sample surface, the samples were sputter coated with While low leachate pHs were observed in UMTpH2.0 and UMTpH3.0,
platinum (Pt) using an ion-sputtering coater equipment (HITACHI the leachate pHs of UMTpH4.0 and UMTpH5.0 were overall higher
E1045). The morphology was then recorded by a field emission than the corresponding leachants, indicating the UMT has a
scanning electron microscope (FESEM, SU8010, Hitachi, Japan) at noticeable pH buffering capacity. Such pH buffering capacity is
1.5 kV accelerating voltage and a working distance (WD) of 5.0 mm. usually owing to the presence of natural minerals that have
Meanwhile, the elemental composition of selected spots on the prominent proton adsorption capacity under acidic environments,
specimens were determined by energy dispersive spectrometer including calcite, smectite, allophane, illlite, chlorite, kaolinite,
(EDS). vermiculite and Fe, Al-oxides (Szecsody et al., 2013). Among these
minerals, calcite and chlorite existed in significant quantities in our
2.4.3. TEM/STEM-EDS experiments, based on XRD results (Fig. 2). Compared to chlorite,
Post-leaching UMT powder was dispersed in alcohol for 5 min, the dissolution of calcite typically results in much more significant
and then was deposited onto a carbon-coated copper grid, which increase of leachate pH (Szecsody et al., 2013). Accordingly, the
was left to stand for 10 min prior to being analyzed by a 200 kV JEOL relatively rapid rise of leachate pH from UMTpH4.0, UMTpH5.0 and
JEM-2100 high-resolution transmission electron microscope. The UMTpH6.0 in some leaching periods might be predominantly
distribution and concentrations of targeted elements in represen- attributed to the dissolution of calcite. It is important to note that
tative particles from post-leaching UMT were determined by STEM- the sharp fluctuation of leachate pH until the 13th week and defi-
EDS, and the U-containing minerals were identified using selected- nite pH decrease at the later stage in UMTpH6.0, reflected the
area electron diffraction (SAED) patterns. opposite impacts on pH from the dissolution of alkaline minerals
and the oxidation of sulfide substance, which later prevailed.
2.4.4. FT-IR Dissolved U concentrations in the leachate during the 25-week
Pre- and post-leaching UMT powder were dispersed in KBr by semi-dynamic leaching tests overwhelmingly exceeded 50 mg/L in
compression to form a sheet, and then the FT-IR spectra were UMTpH2.0 and UMTpH3.0 and also occasionally exceeded 50 mg/L in
collected using a Fourier transform infrared spectroscopy (Tensor UMTpH4.0e6.0 (Fig. 1), the limit of Regulations for Radiation and
27, Bruker, Germany) with the wavenumber range of Environmental Protection of Uranium Mining and Metallurgy
400e4000 cm1 and resolution of 0.4 cm1. (GB23727-2009) in China, not to mention the proposed guideline
range initiated by the World Health Organization (WHO) (2e15 mg/
2.4.5. Sequential extraction L). Surprisingly, the leachate U concentrations in UMTpH4.0 were
Sequential extraction was used to determine the chemical lower than those in UMTpH5.0 and close to those in UMTpH6.0, sug-
speciation of U in pre-leaching UMT. The sequential extraction gesting a nonlinear relationship between the extent of U release
procedure was based on a modified BCR procedure recommended and leaching pH.
by IRMM (Institute for Reference Materials and Measurement, The cumulative amount of U release from UMTpH2.0 after 25
Europe). The procedure included three successive steps to distin- weeks was 328.8 mg/kg which exceeded that of UMTpH3.0,
guish different pools of U: (i) a 16 h acetic acid (HOAc, 0.11 M, pH UMTpH4.0, UMTpH5.0 and UMTpH6.0 by a factor of 85, 590, 211 and
 M. Yin et al. / Environmental Pollution 244 (2019) 174e181 177

Fig. 1. Variations of leachate pH and dissolved U concentration versus leaching time.

625, respectively (Fig. 3). Leaching with simulated strong acid rain during leaching. Usually, the lower the pH of leachant, the higher
(pH ¼ 2.0) resulted in the direct dissolution of U-bearing minerals the cumulative U release is expected. However, there was not a
and also the dissolution of some other minerals that suppress U simple correlation between the leaching pH and the U release
adsorption by enhancing ionic strength and changing sorption capability, highlighting the intricacy in mineral-solution
surfaces (Szecsody et al., 2013). In fact, under such strong acid geochemistry of UMT with relatively moderate pH leachants.
condition, U exists in the predominant form of uranyl cation with
strong solubility (Szecsody et al., 2013) and cannot be adsorbed 3.2. Complex leaching behavior controlled by different minerals
(Zachara et al., 2007). UMTpH3.0, UMTpH4.0, UMTpH5.0 and UMTpH6.0
showed a relatively low U concentration in the leachate, which The XRD patterns of UMT before and after leaching were
could be in part attributed to a decrease of U solubility under measured and compared, which is crucial for the interpretation of
leachate pH ranging from 3.18 to 7.29. Moreover, the subsequent U release. The predominant mineral in UMT is quartz (SiO2), fol-
precipitation of non-U minerals, e.g., strengite, dolomite, diaspore lowed by clinochlore (Mg5Al2Si3O10(OH)8), lepidolite (KMg3Al-
and ferrihydrite would be feasible, which could coat the mineral Si3O10OHF), and albite (NaAlSi3O8). It is worth noting that
surfaces and retard U mobilization (Szecsody et al., 2013). clinochlore existed in considerable amounts in both pre- and post-
Notably, the UMT displayed sustained U release in the leaching leaching UMT, while the comparatively high amount of calcite was
test. The cumulative amount (qt) of extractable U versus leaching only observed in pre-leaching UMT. Calcite was expected to play a
time increased rapidly in UMTpH2.0 and UMTpH3.0 while that in vital role in the migration of U from UMT (Kipp et al., 2009), since
UMTpH4.0 and UMTpH6.0 increased slowly, as shown in Fig. 3. the dissolution of calcite due to acidification might lead to libera-
UMTpH5.0 basically showed the same trend as UMTpH4.0 except for tion of soluble uranyl ion, owing to the formation of solution
the 14th and 15th weeks, during which an abrupt U release phe- complexes between uranyl ion and anions such as carbonate,
nomenon was observed. The abrupt U release might be attributed phosphate and sulfate under acidic conditions. The absence of
to the stronger heterogeneity of natural UMT compared to ground calcite in post-leaching UMT was corroborated from XRD patterns,
tailings and the unsystematic exposure of U-containing minerals which might have contributed to the release of carbonates-bound U
178 M. Yin et al. / Environmental Pollution 244 (2019) 174e181

Fig. 2. XRD patterns of pre- and post-leaching UMT. B(104) denotes calcite, which was
Fig. 4. FT-IR spectra of pre- and post-leaching UMT. Individual spectra are vertically
only observed in pre-leaching UMT. D(001) and D(102) are attributed to meta-autunite,
offset for clarity.
a secondary mineral, which were not observed in pre-leaching UMT and negligible in
UMTpH2.0.

SEM-EDS (Fig. S1) provided additional important information

on the variations of morphology and elemental composition in the
pre- and post-leaching UMT. A significant amount of CaCO3 was
found in pre-leaching UMT while was not observed in post-
leaching UMT, which was in accordance with the XRD results.
Provided the findings from SEM-EDS and XRD, more detailed and
interesting features of the interactions between UMT and leachants
were identified. The exhibited compact structure of pre-leaching
UMT appeared to be different from the morphology of post-
leaching UMT, indicating the corrosion of silicate after being
leached and weathered for prolonged period. In addition, the
concomitant increase of specific surface area may increase the rate
of mineral dissolution but may also promote the re-adsorption of U
on the mineral surface during solid-liquid equilibrium.
The FT-IR spectra of UMT before and after leaching help us to
further understand the leaching mechanism (Fig. 4). A broad band
at about 1435 cm1 was presented in pre-leaching UMT, which can
be assigned to carbonate ions, corresponding to calcite (Valente
et al., 2012). This broad band was merely observed at pre-
leaching UMT, confirming the complete dissolution of calcite after
Fig. 3. The cumulative amount (qt) of extractable U versus leaching time. leaching, in good agreement with XRD and SEM-EDS results.
Moreover, a shoulder peak at ~777 cm1 was presented in all UMT,
almost with no pH dependent spectral changes. This shoulder peak
during the leaching test. Clay minerals, such as clinochlore, often could be attributed to (UO2)2þ stretching vibration from urano-
have a high adsorption capacity for U, which however is 
phane (Ca[UO2]2[SiO3(OH)]2$5H2O) (Cejka, 1999). The weak peaks
exchangeable (Kim et al., 2016). Sequential extractions showed that also from all UMT at around 1640 cm1 were assigned to P-O and/or
pre-leaching UMT possessed remarkably high contents of H-O-H (Günister et al., 2007; Zhou et al., 2014). The co-occurrence
exchangeable and acid-extractable U (60.46%, shown in Table 1). of stronger peaks/bands of P-O and/or H-O-H, (UO2)2þ and Si-O in
The reactions related to calcite and clinochlore could be deemed as the post-leaching UMT, relative to pre-leaching, suggested the
the predominant mechanism of U release in the presence of possible formation of uranium-silicate/phosphate phases after 25
simulated acid rain. It's also worth noting that, a U-bearing sec- weeks of semi-dynamic leaching. These U-bearing secondary
ondary mineral, i.e., meta-autunite (Ca(UO2)2(PO4)2$3H2O) was minerals could form under our experimental condition at room
observed merely in post-leaching UMT, suggesting the precipita- temperature (Nguyen et al., 1992). Such formation might also be
tion of new U-containing minerals only after UMT suffered from partly the reason why the release of U from UMT slowed down at
long-time intermittent leaching and weathering. the later leaching stages.

Table 1
The content (average ± standard deviation) in mg kg1 and portion (%) of U in specific geochemical fraction in the pre-leaching uranium mill tailings (UMT).

Geochemical fraction Acid-extractable Reducible Oxidizable Residual

U content (portion) 617.2 ± 3.1 (60.46%) 152.35 ± 3.5 (14.92%) 221.62 ± 4.2 (21.71%) 29.68 ± 1.9 (2.91%)
 M. Yin et al. / Environmental Pollution 244 (2019) 174e181 179

Fig. 5. TEM image and SAED pattern of a representative particle in post-leaching UMTpH6.0, a) TEM image. b) SAED pattern of the [4 08] zone axis of autunite (Ca(UO2)2(-
PO4)2$11H2O) obtained from the red rectangle of a). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

TEM/STEM-EDS characterization was additionally conducted to slope of the linear regression of the logarithm of cumulative flux
identify the U-retaining phases in the example of UMTpH6.0., shown (Bac i) versus time in logarithmic scale (Moon and Dermatas, 2006;
in Fig. S2 and Fig. 5. Clearly, U was concentrated on a particle Torras et al., 2011). As shown in Fig. 6, the leaching patterns for
(prolate spheroid shape) while P and Ca were densely distributed UMTpH2.0, UMTpH3.0 and UMTpH5.0 were dominated by mineral
on the same particle (Fig. S2). A selected-area electron diffraction dissolution in the beginning but turned to diffusion in the latter
(SAED) on the edge of this particle was conducted to unravel the period. In other words, “shift event” of leaching mechanism
phases that host U. The diffraction spots in the SAED pattern were occurred. In contrast, UMTpH4.0 and UMTpH6.0, both of which
obtained by the cast of [4 08] zone axis of autunite (Ca(UO2)2(- showed relatively low and weekly steady release of U, were gov-
PO4)2$11H2O), with d-spacing of ~10.36 Å and ~2.41 Å (Fig. 5b). erned by the same leaching mechanism during the whole leaching
Consistently, meta-autunite (Ca(UO2)2(PO4)2$3H2O), the product of period. The mechanism was identified to be dissolution and diffu-
autunite dehydration which emerged rapidly in semi-dynamic sion in UMTpH4.0 and UMTpH6.0, respectively.
leaching condition (Sowder et al., 2000), was also identified by Dissolved U concentrations in leachate from UMTpH2.0, UMTpH3.0
XRD in post-leaching UMT. The morphology and size of autunite
particle were illustrated in the TEM image (Fig. 5a). The autunite
particle was similar to a prolate spheroid with a length of ~1 mm, in
which case U migration may be retarded by the effect of capillary
barriers (Um et al., 2010). A previous spectroscopic study on the
mechanism responsible for U release from sediments found that
micrometer-sized metatorbernite (Cu(UO2)2(PO4)2$8H2O) and
more soluble uranophane (Ca[UO2]2[SiO3(OH)]2$5H2O) were the U
minerals that predominantly controlled U release during the in-
teractions between waste discharge and sediments (Arai et al.,
2007). Furthermore, the rapid precipitation of meta-ankoleite
(K(UO2) (PO4)$3H2O) was previously identified when sediments
reacted with acidic U-bearing wastewater, and meta-ankoleite
could easily be entrapped in sediments (Wang et al., 2017a). Like-
wise, autunite (Ca(UO2)2(PO4)2$11H2O) found in UMT in the pre-
sent study could be a result of the long-term interaction between
UMT and simulated acid rain, the formation of which is of signifi-
cant importance for U immobilization.

3.3. Mechanism of uranium release revealed by the leaching

kinetics

Fig. 6. Plot of log Bac i versus time to uncover U release mechanism from UMTpH2.0,
The data from the performed semi-dynamic leaching test could UMTpH3.0, UMTpH4.0, UMTpH5.0 and UMTpH6.0. The identification of release mechanism
be well described by the Fick theory model, which allows the was showed by linear fit of log Bac i versus time, where solid line represents surface
identification of the dominant leaching mechanism based on the dissolution and dash line represents diffusion transport.
180 M. Yin et al. / Environmental Pollution 244 (2019) 174e181

and UMTpH5.0 were overall significantly higher than those from Appendix A. Supplementary data
UMTpH4.0 and UMTpH6.0. The quick dissolution process triggered
fast U release in the beginning stage of the leaching experiment, Supplementary data to this article can be found online at
and lead to the formation of lots of microcrack on the UMT surfaces, https://doi.org/10.1016/j.envpol.2018.10.018.
as evidenced by the SEM images (Fig. S1). Therefore, it is possible
that the rest of U reached saturation in porewaters inside the 5. Notes
microcrack, thereafter U release would be dominated by a diffusion
process. In addition, the diffusion of U from the inner porewaters The authors declare no competing financial interest.
was slower than the mineral dissolution at the leachant-UMT
interface (Elkamash et al., 2006), which explains the slower U
release in the latter leaching period. References

Abdelouas, A., 2006. Uranium mill tailings: geochemistry, mineralogy, and envi-
4. Conclusions ronmental impact. Elements 2, 335e341.
Arai, Y., Marcus, M.A., Tamura, N., Davis, J.A., Zachara, J.M., 2007. Spectroscopic
In this study, the characteristics and mechanism of U release evidence for uranium bearing precipitates in vadose zone sediments at the
Hanford 300-area site. Environ. Sci. Technol. 41, 4633e4639.
from natural U mill tailings (UMT) under simulated rain were un- Carter, M.L., Li, H., Zhang, Y., Vance, E.R., Mitchell, D.R.G., 2009. Titanate ceramics for
veiled by performing semi-dynamic leaching tests for 25 weeks and immobilisation of uranium-rich radioactive wastes arising from 99Mo pro-
conducting complementary mineralogical analyses and modeling. duction. J. Nucl. Mater. 384, 322e326.

Cejka, J., 1999. Infrared spectroscopy and thermal analysis of the uranyl minerals.
The results showed clear differences in the rates and magnitudes of Rev. Mineral. Geochem. 38, 520e622.
U release from UMT under different simulated acid/normal rain. Dahlkamp, F.J., 2009. China, people's republic of. In: Uranium Deposits of the World.
The extent of U release from UMTpH2.0 was significantly higher than Springer, Berlin, Heidelberg, pp. 1e26.
Elkamash, A.M., Elnaggar, M.R., Eldessouky, M.I., 2006. Immobilization of cesium
the other tests, and the nonlinear relationship between leaching pH and strontium radionuclides in zeolite-cement blends. J. Hazard Mater. 136,
and the extent of U release was observed. The presence and 310e316.
dissolution of calcite and clinochlore in UMT are of crucial impor- Fo€rstner, U., 1987. Metal Speciation in Solid Wastes d Factors Affecting Mobility,
pp. 11e41.
tance for explaining different U release behaviors under varied Fang, J., Chi, K.M.L., Lu, Z., Wu, W., 2018. Estimating peak uranium production in
simulated acid/normal rain. Furthermore, autunite (Ca(UO2)2(- China e based on a stella model. Energy Pol. 120, 250e258.
PO4)2$11H2O) might form after intermittent leaching and weath- Ferry, C., Richon, P., Beneito, A., Robe , M.C., 2002. Evaluation of the effect of a cover
layer on radon exhalation from uranium mill tailings: transient radon flux
ering and regulated the further release of U in UMTpH3.0e6.0. Fitting
analysis. J. Environ. Radioact. 63, 49e64.
our experimental data with Fick theory model indicated that, U Günister, E., Pestreli, D., Ünlü, C.H., Atıcı, O., Güngo €r, N., 2007. Synthesis and char-
release was initially controlled by mineral dissolution and then by acterization of chitosan-MMT biocomposite systems. Carbohydr. Polym. 67,
diffusion transport for UMTpH2.0, UMTpH3.0 and UMTpH5.0, whereas 358e365.
Gu, P., Zhang, S., Li, X., Wang, X., Wen, T., Jehan, R., Alsaedi, A., Hayat, T., Wang, X.,
dissolution and diffusion were the univocal release mechanism for 2018. Recent advances in layered double hydroxide-based nanomaterials for
UMTpH4.0 and UMTpH6.0, respectively. the removal of radionuclides from aqueous solution. Environ. Pollut. 240,
This study highlights the necessity of carrying out long-term 493e505.
Han, B., Zhang, E., Cheng, G., Zhang, L., Wang, D., Wang, X., 2018. Hydrothermal
leaching tests to fully understand the mechanism of U release carbon superstructures enriched with carboxyl groups for highly efficient
from UMT, because short-term leaching tests may not reveal the uranium removal. Chem. Eng. J. 338, 734e744.
full picture due to shifts of kinetics influenced by geochemical Hu, R.Z., Bi, X.W., Zhou, M.F., Peng, J.T., Su, W.C., Liu, S., Qi, H.W., 2008. Uranium
metallogenesis in South China and its relationship to crustal extension during
weathering. Moreover, comprehending mineral composition of the cretaceous to tertiary. Econ. Geol. 103, 583e598.
UMT has also shown to be important. For the studied UMT, U is Hu, R.Z., Burnard, P.G., Bi, X.W., Zhou, M.F., Peng, J.T., Su, W.C., Zhao, J.H., 2009.
predominantly bond to exchangeable and acid-extractable fraction, Mantle-derived gaseous components in ore-forming fluids of the Xiangshan
uranium deposit, Jiangxi province, China: evidence from He, Ar and C isotopes.
which has high release potential and therefore requires different
Chem. Geol. 266, 86e95.
management strategy from other UMT in which U is mainly Ilankoon, I.M.S.K., Tang, Y., Ghorbani, Y., Northey, S., Yellishetty, M., Deng, X.,
entrapped in the residual, less reactive fraction. This study also Mcbride, D., 2018. The current state and future directions of percolation
leaching in the Chinese mining industry: challenges and opportunities. Miner.
suggested that the formation of U-bearing secondary mineral
Eng. 125, 206e222.
phases, e.g., autunite (Ca(UO2)2(PO4)2$11H2O) is crucial for U Kim, E., Ahn, H., Jo, H.Y., Ryu, J.H., Koh, Y.K., 2016. Chlorite alteration in aqueous
immobilization when the UMT are exposed to intermittent leaching solutions and uranium removal by altered chlorite. J. Hazard Mater. 327, 161.
and weathering. Kipp, G.G., Stone, J.J., Stetler, L.D., Thompson, A., Vaughan, D.J., 2009. Arsenic and
uranium transport in sediments near abandoned uranium mines in Harding
County, South Dakota. Appl. Geochem. 24, 2246e2255.
Acknowledgements Kosson, D.S., 2002. An integrated framework for evaluating leaching in waste
management and utilization of secondary materials. Environ. Eng. Sci. 19,
159e204.
This work was jointly supported by the National Natural Science Landa, E.R., 2003. Mobilization of radionuclides from uranium mill tailings and
Foundation of China (No.41773011; 41501232; 41573008; related waste materials in anaerobic environments. J. Radioanal. Nucl. Chem.
41873015; 41573119; U1612442), the Natural Science Foundation of 255, 559e563.
Landa, E.R., 2004. Uranium mill tailings: nuclear waste and natural laboratory for
Guangdong Province, China (2017A030313247; 2015A030313512), geochemical and radioecological investigations. J. Environ. Radioact. 77, 1e27.
the Environmental Protection Ministry of Public Welfare Research Liu, J., Luo, X., Wang, J., Xiao, T., Yin, M., Belshaw, N.S., Lippold, H., Kong, L., Xiao, E.,
Projects (201509051), Public Welfare Research Project of Zhejiang Bao, Z.A., 2017. Provenance of uranium in a sediment core from a natural
reservoir, South China: application of Pb stable isotope analysis. Chemosphere
Province, China (LGF18E080013), and the Guangzhou University's 193, 1172e1180.
2017 training program for young top-notch personnel (BJ201709), Liu, J., Wang, J., Li, H., Shen, C.C., Chen, Y., Wang, C., Ye, H., Long, J., Song, G., Wu, Y.,
and High Level University Construction Project of Guangdong 2015. Surface sediment contamination by uranium mining/milling activities in
South China. Clean. - Soil, Air, Water 43, 414e420.
Province (Regional Water Environment Safety and Water Ecological
Luo, J.C., Hu, R.Z., Fayek, M., Bi, X.W., Shi, S.H., Chen, Y.W., 2017. Newly discovered
Protection). The Sponsorship from the China Scholarship Council uranium mineralization at ~2.0 Ma in the Menggongjie granite-hosted uranium
(University of Oxford) granted to Jin Wang was acknowledged. The deposit, South China. J. Asian Earth Sci. 241e249.
help from Dr. Jianying Qi from South China Institute of Environ- Luo, J.C., Hu, R.Z., Fayek, M., Li, C.S., Bi, X.W., Abdu, Y., Chen, Y.W., 2015. In-situ SIMS
uraninite UePb dating and genesis of the Xianshi granite-hosted uranium de-
mental Sciences, Ministry of Ecology and Environmental with posit, South China. Ore Geol. Rev. 65, 968e978.
uranium measurement was highly appreciated. Mason, C., Turney, W., Thomson, B., Lu, N., Longmire, P., Chisholmbrause, C., 1997.
 M. Yin et al. / Environmental Pollution 244 (2019) 174e181 181

Carbonate leaching of uranium from contaminated soils. Environ. Sci. Technol. Preliminary results of spatial distribution of uranium and thorium in soil pro-
31, 2707e2711. files near a uranium industrial site, Guangdong province, China. Nukleonika 61,
Moon, D.H., Dermatas, D., 2006. An evaluation of lead leachability from stabilized/ 367e371.
solidified soils under modified semi-dynamic leaching conditions. Eng. Geol. Wang, J., Liu, J., Li, H., Chen, Y., Xiao, T., Song, G., Chen, D., Wang, C., 2017b. Uranium
85, 67e74. and thorium leachability in contaminated stream sediments from a uranium
Nguyen, S.N., Silva, R.J., Weed, H.C., Jr, J.E.A., 1992. Standard Gibbs free energies of minesite. J. Geochem. Explor. 176, 85e90.
formation at the temperature 303.15 K of four uranyl silicates: soddyite, ura- Wang, J., Liu, J., Li, H., Song, G., Chen, Y., Xiao, T., Qi, J., Zhu, L., 2012a. Surface water
nophane, sodium boltwoodite, and sodium weeksite. J. Chem. Thermodyn. 24, contamination by uranium mining/milling activities in northern Guangdong
359e376. province, China. Acta Hydrochim. Hydrobiol. 40, 1357e1363.
Pan, Z., Giammar, D.E., Mehta, V., Troyer, L.D., Catalano, J.G., Wang, Z., 2016. Phos- Wang, J., Liu, J., Zhu, L., Qi, J.Y., Chen, Y.H., Xiao, T.F., Fu, S.M., Wang, C.L., Li, J.W.,
phate-induced immobilization of uranium in hanford sediments. Environ. Sci. 2012b. Uranium and thorium leached from uranium mill tailing of Guangdong
Technol. 50, 13486e13494. Province, China and its implication for radiological risk. Radiat. Protect. Dosim.
Robinson, P., 2004. Uranium Mill Tailings Remediation Performed by the US DOE: an 152, 215.
Overview. Southwest Research and Information Center, Albuquerque, NM. Waugh, W.J., Kastens, M.K., Sheader, L.R.L., Benson, C.H., Albright, W.H.,
Rufyikiri, G., Wannijn, J., Wang, L., Thiry, Y., 2006. Effects of phosphorus fertilization Mushovic, P.S., 2008. Monitoring the Performance of an Alternative Landfill
on the availability and uptake of uranium and nutrients by plants grown on soil Cover at the Monticello, Utah, Uranium Mill Tailings Disposal Site, Paper Pre-
derived from uranium mining debris. Environ. Pollut. 141, 420. sented at: the Waste Management 2008; Conference, Phoenix 2008 February
Sowder, A.G., Clark, S.B., Fjeld, R.A., 2000. Dehydration of synthetic autunite hy- 24-28.
drates. International Journal for Chemical Aspects of Nuclear Science & Tech- Xu, H., Wang, Y., Zhao, P., Bourcier, W.L., Van, K.R., Shaw, H.F., 2004. Investigation of
nology 88, 533e538. pyrochlore-based U-bearing ceramic nuclear waste: uranium leaching test and
Steinhausler, F., Zaitseva, 2007. Uranium mining and milling: material security and TEM observation. Environ. Sci. Technol. 38, 1480.
risk assessment. Int. J. Nucl. Govern. Econ. Ecol 1, 286e304. Yin, L., Wang, P., Wen, T., Yu, S., Wang, X., Hayat, T., Alsaedi, A., Wang, X., 2017.
Szecsody, J.E., Truex, M.J., Qafoku, N.P., Wellman, D.M., Tom, R., Lirong, Z., 2013. Synthesis of layered titanate nanowires at low temperature and their applica-
Influence of acidic and alkaline waste solution properties on uranium migration tion in efficient removal of U(VI). Environ. Pollut. 226, 125e134.
in subsurface sediments. J. Contam. Hydrol. 151, 155e175. Zachara, J.M., Brown, C.F., Christensen, J.N., Davis, J.A., Dresel, P.E., Liu, C., Kelly, S.D.,
Torras, J., Buj, I., Rovira, M., De, P.J., 2011. Semi-dynamic leaching tests of nickel Mckinley, J.P., Serne, R.J., Um, W., 2007. A Site Wide Perspective on Uranium
containing wastes stabilized/solidified with magnesium potassium phosphate Geochemistry at the Hanford Site. Office of Scientific & Technical Information
cements. J. Hazard Mater. 186, 1954e1960. Technical Reports.
Um, W., Icenhower, J.P., Brown, C.F., Serne, R.J., Wang, Z., Dodge, C.J., Francis, A.J., Zhang, B., Zhang, X.W., Mi, L.I., Tang, D.S., Jun-Wen, L.V., Tan, W.F., 2015. The
2010. Characterization of uranium-contaminated sediments from beneath a Characteristics and Research Progress of Uranium Tailings Comprehensive
nuclear waste storage tank from Hanford, Washington: implications for Treatment Techniques. China Mining Magazine, pp. 58e62 in Chinese.
contaminant transport and fate. Geochem. Cosmochim. Acta 74, 1363e1380. Zhang, C., Liu, Y., Li, X., Chen, H., Wen, T., Jiang, Z., Ai, Y., Sun, Y., Hayat, T., Wang, X.,
Valente, J.S., Rodriguez-Gattorno, G., Valle-Orta, M., Torres-Garcia, E., 2012. Thermal 2018. Highly uranium elimination by crab shells-derived porous graphitic car-
decomposition kinetics of MgAl layered double hydroxides. Mater. Chem. Phys. bon nitride: batch, EXAFS and theoretical calculations. Chem. Eng. J. 406e415.
133, 621e629. Zhang, W., Wang, J., 2017. Leaching performance of uranium from the cement so-
Wang, G., Um, W., Wang, Z., Reinosomaset, E., Washton, N.M., Mueller, K.T., lidified matrices containing spent radioactive organic solvent. Ann. Nucl. En-
Perdrial, N., O'Day, P.A., Chorover, J., 2017a. Uranium release from acidic ergy 101, 31e35.
weathered hanford sediments: single-pass flow-through and column experi- Zhou, L., Jia, Y., Peng, J., Liu, Z., Al-Zaini, E., 2014. Competitive adsorption of ura-
ments. Environ. Sci. Technol. 51, 11011e11019. nium(VI) and thorium(IV) ions from aqueous solution using triphosphate-
Wang, J., Liu, J., Chen, Y., Song, G., Chen, D., Xiao, T., Li, H., Wang, C., Jiang, F., 2016. crosslinked magnetic chitosan resins. J. Radioanal. Nucl. Chem. 302, 331e340.