Journal of Cleaner Production 214 (2019) 860e872

Contents lists available at ScienceDirect

Journal of Cleaner Production

journal homepage: www.elsevier.com/locate/jclepro

Environmentally friendly extraction of gold from refractory

concentrate using a copper e ethylenediamine e thiosulfate solution
Qiang Wang a, Xianzhi Hu b, *, Futing Zi b, **, Peng Yang b, Yunlong Chen a, Shuliang Chen b
a
Faculty of Land Resource Engineering, Kunming University of Science and Technology, Kunming, 650093, China
b
Faculty of Science, Kunming University of Science and Technology, Kunming, 650500, China

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

Article history: The growing environmental concerns associated with the use of cyanides and their toxicity have led to
Received 3 March 2018 the prohibition of cyanidation-based gold leaching in many countries, necessitating the search for other
Received in revised form extraction methods. Herein, we show that thiosulfate leaching presents an alternative green way of gold
6 November 2018
extraction from S- and As-rich refractory gold concentrates, revealing the importance of oxidative ore
Accepted 2 January 2019
Available online 4 January 2019
pretreatment. Prior to leaching, samples were oxidized by roasting, with the optimum roasting tem-
perature (642
C) and time (240 min) determined using the response surface method. Subsequently, the
roasted residues were subjected to cyanidation, copper e ammonia e thiosulfate system, and copper e
Keywords:
Roasting
ethylenediamine e thiosulfate system treatment. As a result, 18-h cyanidation achieved a gold extraction
Refractory gold concentrate efficiency of 80.6%, with the consumption of NaCN equaling 1.05 kg/t, whereas the corresponding values
Response surface method for 1.5 h copper e ammonia e thiosulfate treatment equaled 63.0% and 11.3 kg/t. Importantly, 4 h copper
Ethylenediamine e ethylenediamine e thiosulfate system treatment (5 mM copper ion, 10 mM ethylenediamine, 0.1 M
Thiosulfate thiosulfate, pH ¼ 8e11, liquid-to-solid ratio ¼ 3, stirring speed ¼ 250 rpm) achieved a gold extraction
efficiency of 80.3%, with the corresponding thiosulfate consumption equaling only 4.14 kg/t. Thus, copper
e ethylenediamine e thiosulfate leaching proved to be an efficient, environmentally friendly, and low-
cost method, being well-suited for gold extraction from roasted residues.
© 2019 Elsevier Ltd. All rights reserved.

1. Introduction
The efficiency of gold extraction from sulphidic refractory ores
(e.g., arsenopyrite (FeAsS) and pyrite (FeS2)) by direct cyanidation is
usually very low (Badri and Zamankhan, 2013; Faraz et al., 2014),
since fine gold particles in these ores are mostly locked up in the
bulk and therefore inaccessible to leaching agents and dissolved
oxygen (Badri and Zamankhan, 2013). Moreover, cyanidation of
sulphidic ores consumes a large amount of cyanide and oxygen,
leading to insufficient levels of these species in the pulp during gold
leaching (Dunn and Chamberlain, 1997; Nanthakumar et al., 2007).
To increase the efficiency of gold extraction from sulphidic ores,
they need to be oxidatively pretreated to liberate gold particles
Abbreviations: Response surface methodology, RSM; X-ray diffraction, XRD;
scanning electron microscopy, SEM; energy-dispersive X-ray spectroscopy, EDS;
liquid-to-solid ratio, L/S; central composite design, CCD; atomic absorption spec-
trometry, AAS; inductively coupled plasma-atomic emission spectroscopy, ICP-AES.
* Corresponding author.
** Corresponding author.
E-mail addresses: xianzhihu2@sina.com (X. Hu), 345992103@qq.com (F. Zi).
prior to any conventional leaching treatment, with the main pre-
treatment methods for refractory gold ores being biological
oxidation (Guo et al., 2017; Mubarok et al., 2016; Roberto, 2016),
chemical oxidation (Alp et al., 2014; Nunan et al., 2017), pressure
oxidation (Chan et al., 2015), and roasting oxidation (Michelis et al.,
2013). However, most of these approaches exhibit certain draw-
backs. The environmentally benign biological oxidation is overly
slow. N. V. Fomchenko et al. have reported a two-step biological
oxidation process of gold e bearing sulfide concentrates
(Fomchenko et al., 2016). Yet this environmentally benign process is
overly slow, needing 8 days to reach an extraction efficiency of
94.1%. Chemical oxidation, on the other hand, often incurs high
reagent and maintenance costs. For example, Hasab et al. achieved
82% gold extraction efficiency from 200 g/L pulp concentrate by
using 200 g/L Ca(OCl)2 and 200 g/L NaCl at an initial pH of 11 (Hasab
et al., 2013). The calculated cost of leaching agents is very high.
Pressure oxidation is technically difficult to implement. In their
pressure oxidation leaching of carbonic refractory gold ore, Zhang
et al. reported a sulfur oxidation ratio of 99.35% at the conditions of
210
C and oxygen partial pressure of 0.8 MPa (Zhang et al., 2016).

https://doi.org/10.1016/j.jclepro.2019.01.007
0959-6526/© 2019 Elsevier Ltd. All rights reserved.
 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872
The high temperature and high pressure of this pretreatment
process place high requirements on the equipment, resulting in
high investment costs, rigorous operational technology re-
quirements, and high process costs. Thus, due to its high efficiency,
technical maturity, and excellent reliability, roasting is still the
most commonly used method of refractory gold ore pretreatment
(Bas et al., 2016).
As a mature, efficient, and reliable method of leaching gold,
cyanidation was once widespread, supporting some 90% of the
global gold production. The first recorded use of cyanide to extract
gold from ores was more than a hundred years ago. Gold can be
dissolved in cyanide solutions according to the following reaction:
4Au þ 8CN þ O2 þ 2H2 O ¼ 4AuðCNÞ
2 þ 4OH

(1)
However, even relatively low concentrations of cyanide are
highly toxic to people and wildlife. This drawback led to the pro-
hibition of cyanidation in many countries, making it necessary to
search for other extraction methods.
Thiosulfate-based gold leaching has recently attracted
increasing attention as an alternative to cyanidation (Bisceglie et al.,
2017; Feng and van Deventer, 2010; Ha et al., 2010; Hilson and
Monhemius, 2006; Nie et al., 2016). This method has the advan-
tages of high leaching rate, non-toxicity, cost effectiveness, and
tolerance towards impurities. Moreover, this leaching process is
carried out in a weakly alkaline medium, which is non-corrosive to
the equipment. Although thiosulfate is cheaper than cyanide and
enables faster gold dissolution, the conditions currently reported
for thiosulfate leaching are severe and result in high reagent con-
sumption (Aylmore and Muir, 2001). Recently, ethylenediamine has
been reported as an alternative to ammonia for gold leaching (Nie
et al., 2015; Yu et al., 2014), forming stronger complexes with
copper ions than the latter ligand and therefore reducing the Cu2þ/
Cuþ redox potential. Thus, copper e ethylenediamine e thiosulfate
systems feature copper ions with decreased catalytic oxidation
ability and thus allow the thiosulfate consumption to be lowered.
However, those studies only theoretically analyzed the feasibility of
replacing ammonia with ethylenediamine, while the performance
of this method for carbonaceous refractory gold ore high in sulfur
and arsenic is unknown. Herein, we optimized the roasting of a gold
flotation concentrate for a range of temperatures (400e800
C) and
times (60e300 min) using response surface methodology (RSM)
and subjected the obtained residues to cyanidation, copper e
ammonia e thiosulfate system, and copper e ethylenediamine e
thiosulfate leaching, additionally investigating factors affecting the
gold extraction efficiency.
2. Materials and methods
2.1. Materials and characterization
Gold flotation concentrates obtained from the Yunnan Gold
Mine Group (China) were subjected to particle size screening,
which revealed that 80% of the particles had sizes below 74 mm. In
order to understand the chemical composition of the above con-
centrates in detail, the elemental contents of the ore were deter-
mined by inductively coupled plasma-atomic emission
spectroscopy (ICP-AES, Thermo, USA) and C-S analyzer (LECO, USA),
with the obtained results presented in Table 1. It can be seen from
Table 1
Elemental analysis results of gold flotation concentrates.
Element Au (g/t) Cu Fe S As
Content, % 29.4 0.05 16.7 9.51 1.67
861
Table 1 that the main valuable element for recovery in ore is gold,
which has a grade of 29.4 g/t. In addition, the content of the asso-
ciated harmful elements S and As reached 9.51% and 1.67%,
respectively. In order to achieve the purpose of increasing the gold
extraction efficiency, it is necessary to remove or reduce the con-
tent of harmful elements. The non-metallic minerals of gold con-
centrates were mainly mica and quartz, whereas the metallic
minerals mainly corresponded to fine-grained pyrite and arseno-
pyrite. Knowledge about the occurrence state of gold in ore is the
important to improving the extraction efficiency of gold. Micro-
scopic investigation revealed that the concentrates contained very
fine gold particles, with their maximum size equaling 8.5 mm, and
93.84% of the particles being smaller than 5 mm, which is typical of
refractory sulphidic ores. There is 86.26% of the gold in the sul-
phidic ore, the gold in the gangue is 7.58%, and the free gold only
accounts for 6.16%, implying that these ores are barely amenable to
efficient gold exposure and leaching.
2.2. Methods
2.2.1. Roasting test
A certain quality of gold flotation concentrate was put into a
ceramic crucible. When the furnace reached the preset tempera-
ture, the crucible was put into the furnace quickly, and the timing
was started. After roasting for a specified time, the crucible was
taken out, cooled naturally, and the quality of roasted residue was
weighed. The content of gold in the roasted residue is calculated as
Eq. (2)
m0
b ¼  a (2)
m
where b is the content of gold in the roasted residue (106 g/t), a is
the content of gold in the gold flotation concentrates (106 g/t), m0
is the weight of gold flotation concentrates (g), and m is the weight
of the roasted residue (g).
In the roasting process, the door of the furnace was opened a
crack to permit air flow over the sample. The samples were stirred
by a stainless-steel spatula every 30 min.
2.2.2. Leaching test
The roasted gold concentrate was used for leaching experi-
ments, which were all performed at room temperature (20 ± 2
C)
in a 250 mL Erlenmeyer flask. Leaching solutions were prepared by
mixing chemical agents and distilled water. The pH of the pulp was
adjusted with NaOH or HCl. During leaching, the slurry was
agitated at a constant rate using a mechanical stirrer, and after-
wards it was passed through a pressure filter. The filtrate was
titrated to determine the thiosulfate or cyanide concentration. The
filter cake was thoroughly washed with distilled water, oven-dried
at 100
C for 4 h, and pulverized for gold content analysis. The gold
extraction efficiency (q, %) can be expressed as follows:
mb  m1 g
q ¼  100% (3)
mb
where b is the gold content in the roasted residue (106 g/t), g is the
gold content in the leached residue (106 g/t), m1 is the weight of
leached residue (g), and m is the weight of the roasted residue (g).
C Pb SiO2 MgO Al2O3 CaO
8.29 0.23 42.1 4.67 7.98 5.26
862 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872
2.2.3. Analytical methods
The content of gold in the ore was quantified by atomic ab-
sorption spectrometry (AAS, PerkinElmer, USA). Concentration of
cyanide in the pulp was quantified by titration with silver nitrate,
whereas the concentration of thiosulfate was determined by an
iodometric method. In order to eliminate the effects of SO2 3 and
the cupric ions on iodine titration, a certain amount of formalde-
hyde and ethylenediamine tetraacetic acid, respectively, was added
prior to the titration with the starch indicator. The elemental con-
tents of the roasted residue were determined by ICP-AES and C-S
analyzer. To investigate the microstructure changes of gold flota-
tion concentrates and roasted residue surfaces, the scanning elec-
tron microscopy (SEM, Phenom, Holland) coupled with energy-
dispersive X-ray spectroscopy (EDS) was adopted to characterize
the morphologies. The phase change and crystal structure charac-
teristics of the gold ore before and after roasting were observed by
X-ray powder diffraction (XRD, Rigaku, Japan).
3. Results and discussion
3.1. Comparison test
3.1.1. Gold leaching from non-pretreated refractory ores
In order to determine the refractory nature of the ore, direct
gold leaching from non-pretreated refractory gold concentrates
was conducted by conventional cyanidation and copper e
ammonia e thiosulfate methods.
3.1.1.1. Conventional cyanidation. Refer to the actual production
conditions of the gold mine and studies reports (Bidari and
Aghazadeh, 2018; Ellis and Senanayake, 2004), the gold concen-
trate was leached used the following conditions: liquid-to-solid
ratio (L/S) ¼ 3, stirring speed ¼ 250 rpm, pH ¼ 11 (adjusted with
NaOH), cyanide (NaCN) concentration ¼ 1.0 g/L, and leaching
time ¼ 24 h. After the leaching experiment, the pulp was passed
through a pressure filter, and the filter cake was pulverized for gold
content analysis. The gold extraction efficiency was only 5.9%, and
the cyanide consumption was 1.84 kg/t. Pyrite interferes greatly
with the cyanidation. On the one hand, the gold particles are highly
dispersed and locked up in the sulphidic ores, so they cannot
interact with cyanide during the gold leaching process. On the
other hand, as a major sulphidic ores, pyrite is oxidized in the
slurry, and the oxidation product of ferrous sulfate reacts with cy-
anide to form ferrous cyanate, resulting in large cyanide
consumption:
2FeS2 þ 7O2 þ 2H2 O ¼ 2FeSO4 þ 2H2 SO4
FeSO4 þ 6NaCN ¼ Na4 FeðCNÞ6 þ Na2 SO4
These reactions also lead to an increase in oxygen consumption,
which is not conducive to the process of cyanidation.
3.1.1.2. Copper e ammonia e thiosulfate method. Similarly, in
accordance with the previous study of our laboratory and the
experimental methodology reported in previous studies (Jeffrey,
2001; Liu et al., 2017; Senanayake, 2005), the following parame-
ters were chosen: L/S ¼ 3, stirring speed ¼ 250 rpm, Cu2þ (CuSO4 )
concentration ¼ 5 mM, ammonia (NH3 ,H2 O) concentra-
tion ¼ 0.4 M, thiosulfate (Na2 S2 O3 ) concentration ¼ 0.1 M, pH ¼ 10,
and leaching time ¼ 12 h. The slurry was filtered and the filter cake
was analyzed for the gold content, the extraction efficiency was
even lower (4.3%). The corresponding thiosulfate consumption was
20.4 kg/t. Similarly to cyanidation, the leaching solution could not
contact with gold, resulting in a low extraction efficiency. In
addition, the presence of pyrite catalyzes the decomposition of
thiosulfate with molecular oxygen as an oxidant (Xu and Schoonen,
1995). This results in a large consumption of thiosulfate.
4S2 O2
3 þ O2 þ 4Hþ

! 2S4 O2
6 þ 2H2 O (6)
The above results revealed that the efficiency of direct gold
extraction from S- and As-rich refractory gold ores was very low
due to the gold particles being highly dispersed and locked up in
the ore bulk, hence indicating that these concentrates need to be
pretreated prior to being subjected to conventional leaching
methods.
3.1.2. Roasting oxidation pretreatment test
Roasting was tested as a commonly used pretreatment method.
Specifically, the gold concentrate was roasted for 3 h at 600
C, and
the roasted residue was leached by conventional cyanidation under
the conditions described in section 3.1.1 to achieve a gold extraction
efficiency of 74.3%. Comparison with the above results shows that
roasting is an effective pretreatment method for S- and As-rich
refractory gold ores.
3.2. Optimization of roasting conditions
RSM is a statistical test method to optimize stochastic process.
The purpose of the RSM test design is to obtain a higher precision
response surface model by arranging the test reasonably (Bahrami
et al., 2018; Markandeya et al., 2017). Roasting conditions were
optimized utilizing RSM, and the roasted residue was leached by
conventional cyanidation under the conditions listed in section
3.1.1.
3.2.1. Test design
The efficiency of gold extraction was taken as a response, and
roasting temperature and time were selected as factors of influence,
with their codes and levels used for response surface analysis listed
in Table 2. Central composite design (CCD) is one of the commonly
used experimental design methods in RSM. It uses the limited
number of tests to analyze the influencing factors and their inter-
action, so as to obtain the optimal combination of factors in the
response process. Finally, it estimates a quadratic polynomial
equation in an optimal number of experiments (Çetintaş et al.,
2018; Lu et al., 2017; Lu and Xu, 2017). The experimental setup
and results of CCD test are described in Table 3.
3.2.2. Data analysis
(4) Multiple regression analysis of these results (Table 3) was used
by Design Expert Software. The dependence of gold extraction ef-
(5) ficiency as the evaluation index (dependent variable) on various
factors (independent variables) was expressed as:
Y ¼ e 190.64462 þ 0.62199A þ 0.60241Be0.00039375AB e
0.00041066A2 e 0.000728206B2 (7)
Where Y represents the gold extraction efficiency, and A and B
represent the roasting temperature and time, respectively. The
Table 2
Codes and levels of the two factors used for response surface analysis.
Factor Code Level
1.414 1 0 1 1.414


Temperature ( C) A 317 400 600 800 883
Time (min) B 10 60 180 300 350
 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872 863

Table 3
Experiment design and results of response surface analysis.

Standard order Run A (Temperature/
C) B (Time/min) Y (Gold extraction efficiency/%) Predicted (Gold extraction efficiency/%)

1 6 400 60 19.9 16.5

2 4 800 60 59.8 58.7
3 13 400 300 64.3 60.4
4 2 800 300 66.4 64.8
5 7 317 180 23.5 27.7
6 9 883 180 59.8 60.7
7 8 600 10 36.2 38.4
8 12 600 350 70.8 73.7
9 10 600 180 77.5 77.0
10 3 600 180 78.4 77.0
11 5 600 180 79.0 77.0
12 1 600 180 74.3 77.0
13 11 600 180 75.9 77.0

positive sign indicates the synergistic effect of increasing the gold

extraction efficiency, while the negative sign indicates the antag-
onistic effect.
If the predictive value of the internal residual was a random
distribution, the homogeneity of the residual variance is consistent
with the optimization requirements. Moreover, a normal distribu-
tion of the residuals is an important tool for checking model ac-
curacy. When the residuals conform to the normal probability
distribution, the fitted curve of residuals is linear. Fig. 1 shows the
gold extraction efficiency between the experimental value and the
predicted value. The predicted value is estimated by the model, and
the experimental value of gold extraction efficiency is calculated
from the experimental data. According to Fig. 1, the results were
observed to be a good fit with the predicted values; further, the
fitted curve of experimental and predicted value is linear.
Furthermore, the measured response values and internal residuals
were normally distributed (Fig. 2). In addition, the internal re-
siduals of the gold extraction efficiency model were randomly Fig. 2. Normal probability plots of residuals.
distributed (Fig. 3).
According to Table 4, in the model, A, B, AB, A2 and B2 were
significantly different (P < 0.05). The P values for lack of fit were
found to be higher than 0.05 (Table 4, where the P value of Lack of
Fit was 0.0663), which meant that lack of fit was non-significant in
the gold extraction efficiency model. It indicated that the response
values matched the model, and the model fitted the experimental
data well. In addition, the relationship between predicted and
observed values for models of the gold extraction efficiency was in

Fig. 3. Plot of the residuals versus the predicted response.

good agreement, and R2 value was 0.9853 (Table 4). It suggested

Fig. 1. Predicted gold extraction efficiency versus experimental gold extraction
efficiency.
that the model is adequate and that the model does not show any
violation of the independence or constant variance assumption.
Furthermore, the coefficients of variations (C.V.) were low (5.43%),
which confirmed that the degree of precision of the models and
experimental data were all very high and reliable. Since there is not
much difference between the experimental and predicted values,
the model is acceptable.
864 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872

Table 4
Analysis of response surface variance.

Source Sum of squares df Mean square F-value P-value Significance

***
Model 5067.86 5 1013.57 93.93 <0.0001
***
A 1089.01 1 1089.01 100.93 <0.0001
***
B 1248.25 1 1248.25 115.68 <0.0001
**
AB 357.21 1 357.21 33.11 0.0007
***
A2 1879.76 1 1879.76 174.21 <0.0001
***
B2 768.87 1 768.87 71.26 <0.0001
Residual 75.53 7 10.79
Lack of fit 60.82 3 20.27 5.51 0.0663 no significance
Pure error 14.71 4 3.68
Cor total 5143.39 12

R2 0.9853 C.V.% 5.43

R2Adj 0.9748 RSN 27.114

***:P < 0.0001; **:P < 0.001.

Based on the results listed in Tables 3 and 4, it was concluded

that the utilized RSM could satisfactorily model gold extraction
efficiency.
Fig. 4 and Fig. 5 present a three-dimensional plot and an asso-
ciated contour plot, respectively, showing the effects of A and B on
the gold extraction efficiency. For optimum conditions, these
models provided a visualization method to describe the relation-
ship between response and experimental levels of each factor and
the type of interaction amongst the test variables. The effects of A
(roasting temperature) and B (roasting time) on the gold extraction
efficiency were shown as a 3D-plot and its associated contour plot.
The elliptical shape of the contour plot illustrated that there was a
correlation between A and B. The influence of each factor on the
gold extraction efficiency can be seen intuitively from Fig. 5, and the
steeper curve indicates that the influence of this factor on the gold
extraction efficiency is greater.

Fig. 5. Contour plot of gold extraction efficiency showing interaction between roasting
3.2.3. Optimization and validation of roasting parameters temperature and time at variable levels.
CCD was used to determine optimal roasting temperature/time
(642
C/240 min) and the maximum predicted gold extraction

Fig. 4. Effects of roasting temperature and time on gold extraction efficiency.

 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872 865

efficiency (81.4%). The experimental mean extraction efficiency

determined under optimal conditions equaled 80.8 ± 1.43%, con-
firming the rationality of the RSM model. Thus, Eq. (7) could be
used to accurately predict the efficiency of gold extraction.

3.2.4. Mechanism of roasting

“Roasting” describes the high-temperature oxidative decom-
position of sulfides and arsenic in gold ores. During the calcination
of gold ore, some of its components undergo physical and chemical
changes, causing the metal sulfides to be oxidized into metal oxides
and exposing the wrapped gold. The main reactions in the roasting
process are as follows:

4FeS2 þ 11O2 ¼ 2Fe2 O3 þ 8SO2 (8)

2FeAsS þ 5O2 ¼ Fe2 O3 þ As2 O3 þ 2SO2 (9)

Different refractory gold concentrates have to be roasted at an
appropriate temperature and for an optimal amount of time. An
overly low temperature slows down the oxidation and does not
efficiently expose the gold particles. On the other hand, excessively
high temperatures cause particle sintering and thus result in the
physical encapsulation of gold, decreased porosity of gold
concentrate particles, and a lower gold extraction efficiency.
In addition, SO2 and As2O3 produced during the roasting process
are pollutants. Therefore, when roasting high-sulfur and high-
Fig. 6. SEM images of gold concentrates.
arsenic carbonaceous gold ore, a certain amount of fixative agent
may be added to convert these toxic gases into the corresponding
sulfates and arsenates (Liu et al., 2000; Nazari et al., 2017; Yang
et al., 2013). Commonly used fixatives include calcium oxide, cal-
cium hydroxide, sodium carbonate, sodium hydroxide, and mag-
nesium carbonate.

3.3. Characterization of roasted samples

After roasting, the content of the valuable element Au and the

content of harmful elements S and As in the ore changed. The
elemental contents of the roasted residue obtained under optimum
roasting conditions were determined by ICP-AES and C-S analyzer.
The obtained results are presented in Table 5.
The gold content of the ore increases as the weight of the ore
decreases during the roasting process. A comparison of the
elemental contents of gold concentrate and roasted residue shows
that sulfur and arsenic in the gold concentrate had been effectively
removed. Considering the decrease in the weight of gold ore, the
removal rate of arsenic is calculated to be 73.9%, and that of sulfur is
83.5%.
In order to investigate the impact of roast on gold concentrate,
the morphology and the preliminary elemental composition of gold
concentrate and roasted residue surfaces were conducted by SEM-
EDS analysis. SEM imaging of gold concentrates and the roasted
residue obtained under optimum roasting conditions revealed that
the former particles featured a smooth and dense surface, whereas
those of the latter sample were porous (Fig. 6 and Fig. 8). Moreover,
elemental analysis and SEM - EDS analysis confirmed that roasting
converted the dense sulphidic ore into porous oxidized residues Fig. 7. Corresponding EDS spectra of gold concentrates.
(Fig. 7 and Fig. 9). Hence, roasting can destroy the gold-bearing
sulfide ore and expose the gold. The roasted residue has high

Table 5
Elemental composition of the roasted residue obtained under optimum roasting conditions.

Element Au (g/t) Cu Fe S As C Pb SiO2 MgO Al2O3 CaO

Content, % 35.2 0.05 19.9 2.97 0.33 0.3 0.32 49.3 5.72 9.87 6.84
866 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872

Fig. 10. XRD spectra of (A) pristine gold concentrates and (B) the roasted residue
obtained under optimum roasting conditions.

arsenopyrite originally present in the former were converted into

hematite in the latter.
Thus, elemental, XRD, and SEM-EDS analyses revealed that
roasting destroyed the structure of gold-bearing sulfide ore and
Fig. 8. SEM images of the roasted residue obtained under optimum roasting
greatly increased the particle porosity, exposing the ore-wrapped
conditions. gold to leaching solutions.

3.4. Cyanidation test

Based on the above results, the optimum conditions for gold

concentrate pretreatment were determined as roasting tempera-
ture ¼ 642
C and roasting time ¼ 240 min. The roasted residues
obtained under these conditions were subjected to cyanidation at
NaCN concentrations of 1.0 and 2.0 g/L, pH ¼ 11 (adjusted with
NaOH), L/S ¼ 3, and stirring speed ¼ 250 rpm. As shown in Fig. 11,
the concentration of NaCN had almost no effect on the efficiency of
gold extraction and cyanide consumption. Specifically, the highest
gold extraction efficiency (80.6%) was obtained at a NaCN concen-
tration of 1 g/L after 18 h, with the corresponding cyanide con-
sumption equaling 1.05 kg/t. Thus, the adopted procedure achieved
a satisfactory gold extraction efficiency.
It is well known that cyanide is acutely toxic to humans and
environment (Akcil, 2006; Bisceglie et al., 2017; Johnson, 2015;

Fig. 9. Corresponding EDS spectra of the roasted residue obtained under optimum
roasting conditions.

porosity, large surface area, and chemical inertness. Such a loose

and porous structure and the removal of sulfur and arsenic are
beneficial for the subsequent gold leaching.
Additionally, pristine gold concentrates and roasted residues Fig. 11. Effects of cyanidation time on cyanide consumption and gold extraction
were probed by XRD (Fig. 10), which revealed that the pyrite and efficiency.
 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872
Kumah, 2006). Liquid or gaseous hydrogen cyanide and cyanides
can enter the body by inhalation, ingestion, or absorption through
the eyes and skin. When there is a cut in the skin, the absorption of
hydrogen cyanide or cyanides increases. The lethal dose of
hydrogen cyanide is 1e3 mg per kg of body weight (Kuyucak and
Akcil, 2013; Moussavi and Khosravi, 2010). When the concentra-
tion of gaseous hydrogen cyanide is 100e300 ppm, inhalation of
the air will result in death within 10e60 min, while death will
occur within 1 min when the concentration is 2000 ppm. Even
when at sublethal doses, long-term exposure to these compounds
can cause loss of appetite, headaches and dizziness, optic atrophy,
depressed thyroid function, and damage to the nervous system
(Hilson and Murck, 2001). In addition, if a sufficient amount of
cyanide is accumulated in rivers and lakes, it will threaten the
health of aquatic organisms and cause neurological damage to
animals ingesting it.
Therefore, cyanidation has serious environmental and health
consequences for miners and the community (Vela
squez-Lo

pez
et al., 2011). It is necessary to find a safe, efficient, and environ-
mentally friendly gold leaching method as an alternative.
3.5. Copper e ammonia e thiosulfate leaching test
Thiosulfate, showing considerable promise as an alternative
lixiviant to replace the toxic cyanide (Akcil, 2006; Stenson, 2006),
was herein utilized in copper e ammonia e thiosulfate leaching
(0.1 M Na2S2O3, 5 mM CuSO4, 0.4 M NH3$H2O, pH ¼ 10, L/S ¼ 3, and
stirring speed ¼ 250 rpm). The performance of this method is
illustrated in Fig. 12, which reveals that the extraction efficiency
increased with leaching time, reaching 63.0% after 1.5 h, with the
corresponding thiosulfate consumption equaling 11.3 kg/t ore.
Thiosulfate has proven beneficial as a fertilizer for many decades. It
has distinct advantages over cyanide due to its low toxicity. As a
traditional gold leaching method using thiosulfate, copper e
ammonia e thiosulfate leaching also exhibits several
disadvantages.
On the one hand, the high concentration of ammonia required to
stabilize copper ions and reduce thiosulfate consumption
(Abbruzzese et al., 1995; Aylmore and Muir, 2001; Mohammadi
et al., 2017) causes severe environmental problems and poses a
threat human health (Fugiel et al., 2017; Lei et al., 2009; Rahimpour
and Asgari, 2008). Initially, the leaching slag absorbed a lot of
ammonia after leaching, and the subsequent treatment cost
increased. Furthermore, ammonia is being increasingly recognized
Fig. 12. Effects of copper e ammonia e thiosulfate leaching time on thiosulfate con-
sumption and gold extraction efficiency.
867
as a major atmospheric pollutant, and is considered as a respiratory
irritant for humans at certain concentrations. These air pollutants
can cause direct adverse effects on human health due to over-
exposure (Artín ~ ano et al., 2018). Additionally, the effluent from gold
plants contains an undesirably high amount of ammonia, which
pollutes drinking water and endangers the health of fish and
humans. Finally, ammonia is difficult to decompose and eventually
produces nitrate. The nitrate leads to eutrophication of water and
soils, which shifts the nutrient balance for plant species and may
cause loss of biodiversity (Wichink Kruit et al., 2017). On the other
hand, the overly high thiosulfate consumption is also a drawback
since it can increase the operation cost. In addition, a large amount
of thiosulfate is decomposed into polysulfate, sulfate, and sulfide,
which will complicate the composition of the gold leaching solu-
tion and passivate the dissolution of gold.
3.6. Copper e ethylenediamine e thiosulfate leaching test
Ethylenediamine (en) is an organic amine compound. It has the
typical didentate ligands and can form coordination compounds
with many transition metal ions. In view of the abovementioned
drawbacks of the copper e ammonia e thiosulfate system, ethyl-
enediamine was chosen as an alternative to ammonia in the copper
e ethylenediamine e thiosulfate method.
In the copper e ethylenediamine e thiosulfate leaching system,
the complexing ligands such as ethylenediamine and thiosulfate,
the Cu2þ/Cuþ redox couple, and possible oxidative decomposition
reactions of thiosulfate are simultaneously present. Therefore, the
gold leaching process also becomes very complicated.
A possible mechanism for the thiosulfate leaching reactions is as
follows;
Au þ 5S2 O2
3 þ CuðenÞ2þ
2 ¼ AuðS2 O3 Þ3
2 þ 2en
þ CuðS2 O3 Þ5
3 (10)
4CuðS2 O3 Þ5
3 þ 8en þ O2 þ 2H2 O
¼ 4CuðenÞ2þ
2 þ 4OH þ 12S2 O2
3 (11)
Net Reaction:4Au þ 8S2 O3 þ O2 þ 2H2 O
¼ 4AuðS2 O3 Þ3
2 þ 4OH (12)
In this section, the single factor condition test will be used to
explore the mechanism of gold leaching and thiosulfate con-
sumption during various gold leaching processes, so as to select the
optimal conditions.
3.6.1. Effect of copper ion concentration
During gold leaching, copper acts as an oxidation catalyst,
speeding up the dissolution of gold. Thus, we expected the copper
ion concentration to have a critical effect on gold extraction effi-
ciency, with investigations performed under the following condi-
tions: en/Cu2þ mole ratio ¼ 2, thiosulfate concentration ¼ 0.1 M,
pH ¼ 9, L/S ¼ 5, reaction time ¼ 4 h, and stirring speed ¼ 250 rpm.
As depicted in Fig. 13, the efficiency of gold extraction considerably
increased with increasing copper ion concentration, reaching a
saturation value of 80.1% at [Cu2þ] ¼ 5 mM.
The concentration of free copper ions in the leaching solution
increases with increasing total copper concentration, having a great
influence on the consumption of thiosulfate (Senanayake, 2004), as
confirmed by the trend observed in Fig. 13 (copper ion concentra-
tion was positively correlated with thiosulfate consumption).
In the copper e ethylenediamine e thiosulfate system,
increasing the concentration of copper ions helps increase the
868 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872

Au þ 2S2 O2
3 ¼ AuðS2 O3 Þ3
2 þ e (18)
The mixed potential of the leaching solution can be obtained
from Eq. (19)
h ih i3
2þ 2
0:059 CuðenÞ 2 S 2 O 3
f ¼ fQ þ lg h i (19)
z CuðS2 O3 Þ5 ½en 2
3

h i h i
¼  0:20344 þ 0:059lg CuðenÞ2þ 2 þ 0:177lg S2 O2
3
h i
 0:059lg CuðS2 O3 Þ5
3  0:118lg½en

When the copper ion concentration is constant, an excessively

high ethylenediamine concentration also decreases the mixed po-
tential of the leaching solution. Thus, the dissolution rate of gold is
Fig. 13. Effect of copper ion concentration on gold extraction efficiency and thiosulfate lowered.
consumption.
Gold extraction at variable ethylenediamine concentrations
(5e15 mM) was investigated under the following conditions: cop-
mixed potential of the slurry. Compared with CuðNH3 Þ2þ 2þ per ion concentration ¼ 5 mM, initial thiosulfate concentra-
4 , CuðenÞ2
has a larger volume and a slower diffusion rate. Therefore, tion ¼ 0.1 M, L/S ¼ 5, pH ¼ 9, reaction time ¼ 4 h, and stirring
increasing the concentration of copper ions is beneficial to increase speed ¼ 250 rpm (Fig. 14). The obtained results showed that the
the CuðenÞ2þ efficiency of gold extraction reaches its maximum (80.1%) at an en/
2 concentration on the gold surface, which is beneficial
to gold leaching. However, excessive copper ions strongly catalyze Cu2þ molar ratio of two, decreasing at higher molar ratios.
the decomposition of thiosulfate, which is the main reason for the In the copper e ethylenediamine e thiosulfate system, the
large thiosulfate consumption in the gold leaching process. The concentration of ehylenediamine also greatly influences the con-
main chemical reactions are as follows: sumption of thiosulfate. At low ethylenediamine concentrations,
copper ions were not fully complexed, increasing the consumption
2Cu2þ þ 2S2 O2 ¼ 2Cuþ þ S4 O2 (13) of thiosulfate. At en/Cu2þ molar ratios of greater than two, the
3 6
copper ions were completely complexed and existed mainly in the
form of CuðenÞ2þ 2 , thus preventing the oxidative decomposition of
Cu2þ þ S2 O2
3 þ 2OH ¼ CuS þ SO2
4 þ H2 O (14) thiosulfate. Therefore, the thiosulfate consumption did not signif-
icantly change at ethylenediamine concentrations above 10 mM.
2CuðenÞ2þ
2 þ 8S2 O2
3 ¼ 2CuðS2 O3 Þ5
3 þ S4 O2
6 þ 4en (15) Considering the combined effects of ethylenediamine concentra-
tion on the efficiency of gold extraction and thiosulfate consump-
tion, an optimal level of 10 mM was chosen for leaching.
2CuðS2 O3 Þ5
3 þ 2OH ¼ Cu2 S þ SO2
4 þ 5S2 O2
3 þ H2 O
(16)
3.6.3. Effect of thiosulfate concentration
Therefore, copper ions play both positive and negative roles The effect of thiosulfate concentration (0.02e0.18 M) on gold
during gold leaching, catalyzing the leaching of gold and also extraction was investigated under the following conditions: copper
causing a large consumption of thiosulfate. To reduce the latter, it is ion concentration ¼ 5 mM, ethylenediamine concentra-
necessary to control the copper ion concentration while ensuring tion ¼ 10 mM, L/S ¼ 5, pH ¼ 9, reaction time ¼ 4 h, and stirring
high gold extraction efficiency. For the following experiments, a speed ¼ 250 rpm (Fig. 15). The obtained results revealed that the
copper ion concentration of 5 mM was chosen.

3.6.2. Effect of ethylenediamine concentration

The main role of ethylenediamine in the copper e ethylenedi-
amine e thiosulfate system is to stabilize the copper ions. The
concentration of ethylenediamine should be properly controlled to
ensure the formation of a stable complex with copper ions. In an
alkaline environment, too low ethylenediamine concentration is
insufficient to form a stable complex with cooper ions and some of
the copper ions precipitate as copper hydroxide. The decrease of
copper ions in the solution will lower the mixed potential of the
leaching solution, which in turn reduces the dissolution rate of
gold.
The leaching of gold in thiosulfate solutions is an electro-
chemical reaction. These half-cell reactions of Eq. (10) are shown in
Eqs. (17) and (18), respectively

CuðenÞ2þ
2 þ 3S2 O2
3 þ e ¼ CuðS2 O3 Þ5
3 þ 2en (17)
Fig. 14. Effect of ethylenediamine concentration on gold extraction efficiency and
thiosulfate consumption.
 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872 869

Fig. 15. Effect of thiosulfate concentration on gold extraction efficiency and thiosulfate Fig. 16. Effect of the L/S ratio on gold extraction efficiency and thiosulfate
consumption. consumption.

gold dissolution rate rapidly increased with increasing thiosulfate

concentration, as rationalized by Eq. (12). Kinetically, a higher
thiosulfate concentration increases the forward reaction rate, and 4CuðenÞþ 2þ
2 þ O2 þ 2H2 O ¼ 4CuðenÞ2 þ 4OH (20)
this is conducive to the dissolution and extraction of gold.
However, excess thiosulfate will react with other oxidants (such
as dissolved oxygen and CuðenÞ2þ 2 in the gold leaching solution)
4CuðS2 O3 Þ5
3 þ O2 þ 2H2 O þ 8en
during leaching. Hence, the thiosulfate is easily decomposed to ¼ 4CuðenÞ2þ þ 12S2 O2 þ 4OH (21)
2 3
produce sulfates, sulfides, and polythionate. It increases the con-
sumption of thiosulfates, and the production of polythionate will It can be seen that the high stirring speed is conducive to the
affect the subsequent recovery of gold. From Fig. 15, when the dissolution of gold. Herein, the influence of stirring speed
concentration of thiosulfate is 0.1 M, the gold extraction efficiency (50e450 rpm) on gold extraction from roasted residues was
reaches the maximum. investigated under the following conditions: copper ion concen-
However, since a thiosulfate concentration beyond 0.1 M resul- tration ¼ 5 mM, ethylenediamine concentration ¼ 10 mM, initial
ted in elevated reagent consumption and did not largely impact thiosulfate concentration ¼ 0.1 M, pH ¼ 9, reaction time ¼ 4 h, and
gold extraction, a value of 0.1 M was chosen as the optimal thio- L/S ratio ¼ 3 (Fig. 17). The obtained results showed that the effi-
sulfate concentration. ciency of gold extraction increased with increasing stirring speed
up to 250 rpm, while the saturation observed at higher speeds in-
dicates that leaching was no longer controlled by aqueous-phase
3.6.4. Effect of the L/S ratio transport. In addition, the increasing stirring speed resulted in an
The L/S ratio significantly impacts both leaching speed and ef- increased consumption of thiosulfate, which was ascribed to the
ficiency. The effect of this ratio (1e5) was investigated under the enrichment of the leaching solution with oxygen under intensive
following conditions: copper ion concentration ¼ 5 mM, ethyl- stirring and the resulting thiosulfate scavenging (Feng and van
enediamine concentration ¼ 10 mM, initial thiosulfate concentra- Deventer, 2007; Jeffrey et al., 2003). Thus, the stirring speed was
tion ¼ 0.1 M, pH ¼ 9, reaction time ¼ 4 h, and stirring kept at 250 rpm for all following experiments.
speed ¼ 250 rpm. At small L/S ratios, the high pulp concentration
resulted in poor dispersion of the leaching agent, slowing down
gold extraction. Notably, the gold leaching efficiency increased with
increasing L/S ratio, since the concomitant decrease of pulp vis-
cosity favored the diffusion of the leaching agent and reaction
products. However, operation at increased L/S ratios incurs costs
associated with equipment investment and reagent consumption.
As seen in Fig. 16, the efficiency of gold extraction did not increase
at L/S ratios above three, whereas the consumption of thiosulfate
increased. Based on the above considerations, an L/S ratio of three
was chosen as an optimal value.

3.6.5. Effect of stirring speed

Agitation is known to significantly influence the leaching of
gold, affecting the contact between the leaching liquid and solid
particles and increasing the content of oxygen in the extractant.
When the agitation speed is high, a large amount of oxygen enters
the solution. Oxygen can rapidly oxidize the reduction products
CuðenÞþ 5 2þ
2 and CuðS2 O3 Þ3 to CuðenÞ2 according to the following Fig. 17. Effect of stirring speed on gold extraction efficiency and thiosulfate
reactions: consumption.
870 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872
3.6.6. Effect of leaching time
In general, the gold extraction efficiency will increase with time.
However, prolonged leaching will reduce the equipment handling
capacity, increase the investment costs, and increase the reagent
consumption. Therefore, choosing the appropriate leaching time is
critical. The influence of leaching time on gold extraction was
investigated under the following conditions: copper ion concen-
tration ¼ 5 mM, ethylenediamine concentration ¼ 10 mM, initial
thiosulfate concentration ¼ 0.1 M, pH ¼ 9, L/S ¼ 3, and stirring
speed ¼ 250 rpm (Fig. 18). The obtained results showed that the
gold extraction efficiency increased from 56.7 to 80.3% with
increasing leaching time, reaching saturation at 4 h. The con-
sumption of thiosulfate also increased with leaching time, equaling
4.14 kg/t at 4 h. A leaching time longer than 4 h will not increase the
extraction efficiency of gold, but will increase the consumption of
thiosulfate.
3.6.7. Effect of pH
In the acidic medium, thiosulfate is easily decomposed accord-
ing to the following reaction:
S2 O2
3 þ 2Hþ ¼ SO2 þ S þ H2 O
Leaching gold in an acidic system causes large consumption of
thiosulfate, low gold leaching efficiency, and equipment corrosion.
Therefore, gold leaching using the copper e ethylenediamine e
thiosulfate system must be carried out in an alkaline medium.
The influence of pH (7e12) was investigated under the
following conditions: copper ion concentration ¼ 5 mM, ethyl-
enediamine concentration ¼ 10 mM, thiosulfate concentra-
tion ¼ 0.1 M, reaction time ¼ 4 h, L/S ¼ 3, and stirring
speed ¼ 250 rpm. The high stability constant of CuðenÞ2þ 2 implies
that this complex is stable at pH ¼ 6.0e11.0 in the copper e eth-
ylenediamine e thiosulfate medium. The highest thiosulfate con-
sumption observed at pH ¼ 7.0 was ascribed to the formation of
trace amounts of Fe3þ, which efficiently catalyzes thiosulfate
decomposition. In addition, a small amount of H2en2þ is formed in
the solution, and the product could no longer form CuðenÞ2þ 2 with
Fig. 18. Effect of leaching time on gold extraction efficiency and thiosulfate consumption.
Cu2þ . This causes an increase in free copper ions and increased
consumption of thiosulfate. However, at pH > 11, the partial con-
version of CuðenÞ2þ2 into copper(II) hydroxide decreased the gold
leaching efficiency and thiosulfate consumption. Fig. 19 shows that
the above two parameters remained almost unchanged as the pH
was increased from 8 to 11.
In contrast to CuðenÞ2þ 2þ
2 , CuðNH3 Þ4 is stable in a narrow pH re-
gion of 9.5e10.0 in the copper e ammonia e thiosulfate medium
(Aylmore and Muir, 2001). Thus, compared to copper e ammonia e
thiosulfate leaching, copper e ethylenediamine e thiosulfate
leaching has several obvious advantages, offering more effective
gold extraction and decreased thiosulfate consumption, addition-
ally operating at a very low ligand (en) concentration. Finally,
copper e ethylenediamine e thiosulfate leaching can be performed
within a much wider pH range.
4. Conclusions
This paper mainly studied an environmentally friendly method
of gold extraction from refractory concentrate. First, the gold
concentrate was pretreated by roasting oxidation, and then the
obtained residue was leached separately using cyanidation, copper
(22)
e ammonia e thiosulfate, and copper e ethylenediamine e thio-
sulfate systems. The effects of various factors on the consumption
of reagents and the gold extraction efficiency in the copper e
ethylenediamine e thiosulfate system were discussed in detail, and
the related mechanism was studied. The following conclusions
were drawn.
(1) Roasting oxidation was shown to be an effective method for
the pretreatment of refractory gold concentrate, with the
RSM-predicted optimum conditions corresponding to a
roasting temperature of 642
C and a roasting time of
240 min. Cyanidation of the thus-treated concentrate ach-
ieved a gold extraction efficiency of 80.6% after 18 h, with the
consumption of NaCN equaling 1.05 kg/t.
(2) XRD and SEM analyses revealed that the pyrite and arseno-
pyrite originally present in gold concentrates were
 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872
Fig. 19. Effect of pH on gold extraction efficiency and thiosulfate consumption.
transformed into hematite upon roasting, which resulted in
decomposition of the dense sulphidic ore inclusions and
made the ore particles more loose and porous, exposing the
sulphidic ore-wrapped gold particles to the leaching
solution.
(3) In the case of traditional copper e ammonia e thiosulfate
leaching, the maximal gold extraction efficiency (63.0%) was
observed after 1.5 h, with the corresponding consumption of
thiosulfate equaling 11.3 kg/t. Under optimum conditions, a
much higher gold extraction efficiency of 80.3% was realized
in the copper e ethylenediamine e thiosulfate system, with
the thiosulfate consumption equaling only 4.14 kg/t. Thus,
the newly developed method had an efficiency similar to that
of cyanidation, but with much lower cost and environmental
impact.
(4) Copper e ethylenediamine e thiosulfate leaching can be
used in a wide range of pH. Moreover, the concentration of
ethylenediamine used is much lower than that of ammonia
in the copper e ammonia e thiosulfate leaching, making the
former method more environmentally friendly. Copper e
ethylenediamine e thiosulfate leaching leads to not only to
high gold extraction efficiency but also greatly reduced
consumption of thiosulfate. Therefore, this leaching system
is more suitable for gold extraction from S- and As-rich
carbonaceous refractory gold ore, as compared to cyanida-
tion or copper e ammonia e thiosulfate leaching.
(5) These results will provide a theoretical basis for the estab-
lishment of future gold extraction systems with high effi-
ciency and low thiosulfate consumption. They also provide
experimental basis for the industrialization of this non-toxic
gold leaching technology from refractory ore.
Acknowledgements
This work was financially supported by the Natural Science
Foundation of China (51674128, 21764008).
871
References
Abbruzzese, C., Fornari, P., Massidda, R., Veglio  , F., Ubaldini, S., 1995. Thiosulphate
leaching for gold hydrometallurgy. Hydrometallurgy 39 (1), 265e276.
Akcil, A., 2006. Managing cyanide: health, safety and risk management practices at
Turkey's Ovacik goldesilver mine. J. Clean. Prod. 14 (8), 727e735.
_ Celep, O., Paktun, Ccedil, D., Thibault, Y., 2014. Influence of potassium hy-
Alp, I.,
droxide pretreatment on the extraction of gold and silver from a refractory ore.
Hydrometallurgy 146 (3), 64e71.
Artín~ ano, B., Pujadas, M., Alonso-Blanco, E., Becerril-Valle, M., Coz, E., Go
mez-
Moreno, F.J., Salvador, P., Nun ~ ez, L., Palacios, M., Diaz, E., 2018. Real-time
monitoring of atmospheric ammonia during a pollution episode in Madrid
(Spain). Atmos. Environ. 189, 80e88.
Aylmore, M.G., Muir, D.M., 2001. Thiosulfate leaching of goldda review. Miner. Eng.
14 (2), 135e174.
Badri, R., Zamankhan, P., 2013. Sulphidic refractory gold ore pre-treatment by se-
lective and bulk flotation methods. Adv. Powder Technol. 24 (2), 512e519.
Bahrami, H., Eslami, A., Nabizadeh, R., Mohseni-Bandpi, A., Asadi, A., Sillanp€ €, M.,
aa
2018. Degradation of trichloroethylene by sonophotolytic-activated persulfate
processes: optimization using response surface methodology. J. Clean. Prod.
198, 1210e1218.
Bas, A.D., Safizadeh, F., Ghali, E., Choi, Y., 2016. Leaching and electrochemical
dissolution of gold in the presence of iron oxide minerals associated with
roasted gold ore. Hydrometallurgy 166, 143e153.
Bidari, E., Aghazadeh, V., 2018. Pyrite from Zarshuran Carlin-type gold deposit:
characterization, alkaline oxidation pretreatment, and cyanidation. Hydromet-
allurgy 179, 222e231.
Bisceglie, F., Civati, D., Bonati, B., Faraci, F.D., 2017. Reduction of potassium cyanide
usage in a consolidated industrial process for gold recovery from wastes and
scraps. J. Clean. Prod. 142, 1810e1818.
€l, D., 2018. A novel reagent-assisted mechanochemical
Çetintaş, S., Yildiz, U., Bingo
method for nickel recovery from lateritic ore. J. Clean. Prod. 199, 616e632.
Chan, T., Collins, M., Dennett, J., Stiksma, J., Ji, J., Kalanchey, R., Berezowsky, R., 2015.
Pilot plant pressure oxidation of refractory gold-silver concentrate from
Eldorado Gold Corporation's Certej Project in Romania. Can. Metall. Q. 54 (3),
252e260.
Dunn, J.G., Chamberlain, A.C., 1997. The recovery of gold from refractory arseno-
pyrite concentrates by pyrolysis-oxidation. Miner. Eng. 10 (9), 919e928.
Ellis, S., Senanayake, G., 2004. The effects of dissolved oxygen and cyanide dosage
on gold extraction from a pyrrhotite-rich ore. Hydrometallurgy 72 (1e2),
39e50.
Faraz, S., Hossna, D., Rezgar, B., Piroz, Z., 2014. Improved recovery of a low-grade
refractory gold ore using flotationepreoxidationecyanidation methods. Int. J.
Min. Sci. Technol. 24 (4), 537e542.
Feng, D., van Deventer, J.S.J., 2007. The role of oxygen in thiosulphate leaching of
gold. Hydrometallurgy 85 (2e4), 193e202.
Feng, D., van Deventer, J.S.J., 2010. Oxidative pre-treatment in thiosulphate leaching
of sulphide gold ores. Int. J. Miner. Process. 94 (1e2), 28e34.
Fomchenko, N.V., Kondrat'eva, T.F., Muravyov, M.I., 2016. A new concept of the
872 Q. Wang et al. / Journal of Cleaner Production 214 (2019) 860e872
biohydrometallurgical technology for gold recovery from refractory sulfide
concentrates. Hydrometallurgy 164, 78e82.
Fugiel, A., Burchart-Korol, D., Czaplicka-Kolarz, K., Smolin
ski, A., 2017. Environ-
mental impact and damage categories caused by air pollution emissions from
mining and quarrying sectors of European countries. J. Clean. Prod. 143,
159e168.
Guo, Y., Guo, X., Wu, H., Li, S., Wang, G., Liu, X., Qiu, G., Wang, D., 2017. A novel bio-
oxidation and two-step thiourea leaching method applied to a refractory gold
concentrate. Hydrometallurgy 171, 213e221.
Ha, V.H., Lee, J.C., Jeong, J., Hai, H.T., Jha, M.K., 2010. Thiosulfate leaching of gold
from waste mobile phones. J. Hazard Mater. 178 (1e3), 1115e1119.
Hasab, M.G., Rashchi, F., Raygan, S., 2013. Simultaneous sulfide oxidation and gold
leaching of a refractory gold concentrate by chlorideehypochlorite solution.
Miner. Eng. 50e51, 140e142.
Hilson, G., Monhemius, A.J., 2006. Alternatives to cyanide in the gold mining in-
dustry: what prospects for the future? J. Clean. Prod. 14 (12e13), 1158e1167.
Hilson, G., Murck, B., 2001. Progress toward pollution prevention and waste mini-
mization in the North American gold mining industry. J. Clean. Prod. 9 (5),
405e415.
Jeffrey, M.I., 2001. Kinetic aspects of gold and silver leaching in
ammoniaethiosulfate solutions. Hydrometallurgy 60 (1), 7e16.
Jeffrey, M.I., Breuer, P.L., Chu, C.K., 2003. The importance of controlling oxygen
addition during the thiosulfate leaching of gold ores. Int. J. Miner. Process. 72
(1e4), 323e330.
Johnson, C.A., 2015. The fate of cyanide in leach wastes at gold mines: an envi-
ronmental perspective. Appl. Geochem. 57, 194e205.
Kumah, A., 2006. Sustainability and gold mining in the developing world. J. Clean.
Prod. 14 (3), 315e323.
Kuyucak, N., Akcil, A., 2013. Cyanide and removal options from effluents in gold
mining and metallurgical processes. Miner. Eng. 50e51, 13e29.
Lei, X., Li, M., Zhang, Z., Feng, C., Bai, W., Sugiura, N., 2009. Electrochemical
regeneration of zeolites and the removal of ammonia. J. Hazard Mater. 169
(1e3), 746e750.
Liu, J., Chi, R., Xu, Z., Zeng, Z., Liang, J., 2000. Selective arsenic-fixing roast of re-
fractory gold concentrate. Metall. Mater. Trans. B 31 (6), 1163e1168.
Liu, X., Xu, B., Yang, Y., Li, Q., Jiang, T., Zhang, X., Zhang, Y., 2017. Effect of galena on
thiosulfate leaching of gold. Hydrometallurgy 171, 157e164.
Lu, Y., Song, Q., Xu, Z., 2017. Integrated technology for recovering Au from waste
memory module by chlorination process: selective leaching, extraction, and
distillation. J. Clean. Prod. 161, 30e39.
Lu, Y., Xu, Z., 2017. Recycling non-leaching gold from gold-plated memory cards:
parameters optimization, experimental verification, and mechanism analysis.
J. Clean. Prod. 162, 1518e1526.
Markandeya, Dhiman, N., Shukla, S.P., Kisku, G.C., 2017. Statistical optimization of
process parameters for removal of dyes from wastewater on chitosan ceno-
spheres nanocomposite using response surface methodology. J. Clean. Prod.
149, 597e606.
Michelis, I.D., Olivieri, A., Ubaldini, S., Ferella, F., Beolchini, F., Veglio  , F., 2013.
Roasting and chlorine leaching of gold-bearing refractory concentrate: experi-
mental and process analysis. Int. J. Min. Sci. Technol. 23 (5), 709e715.
Mohammadi, E., Pourabdoli, M., Ghobeiti-Hasab, M., Heidarpour, A., 2017. Ammo-
niacal thiosulfate leaching of refractory oxide gold ore. Int. J. Miner. Process.
164, 6e10.
Moussavi, G., Khosravi, R., 2010. Removal of cyanide from wastewater by adsorption
onto pistachio hull wastes: parametric experiments, kinetics and equilibrium
analysis. J. Hazard Mater. 183 (1e3), 724e730.
Mubarok, M.Z., Winarko, R., Chaerun, S.K., Rizki, I.N., Ichlas, Z.T., 2016. Improving
gold recovery from refractory gold ores through biooxidation using iron-sulfur-
oxidizing/sulfur-oxidizing mixotrophic bacteria. Hydrometallurgy 168, 69e75.
Nanthakumar, B., Pickles, C.A., Kelebek, S., 2007. Microwave pretreatment of a
double refractory gold ore. Miner. Eng. 20 (11), 1109e1119.
Nazari, A.M., Radzinski, R., Ghahreman, A., 2017. Review of arsenic metallurgy:
treatment of arsenical minerals and the immobilization of arsenic. Hydromet-
allurgy 174, 258e281.
Nie, Y., Chi, H., Zi, F., Hu, X., Yu, H., He, S., 2015. The effect of cobalt and nickel ions on
gold dissolution in a thiosulfate-ethylenediamine (en)-Cu2þ system. Miner.
Eng. 83, 205e210.
Nie, Y., Yu, Q., Hu, X., Zi, F., Yu, H., 2016. The effect of ammonia on the anodic process
of gold in copper-free thiosulfate solution. J. Electrochem. Soc. 163 (5),
E123eE129.
Nunan, T.O., Viana, I.L., Peixoto, G.C., Ernesto, H., Verster, D.M., Pereira, J.H.,
Bonfatti, J.M., Teixeira, L.A.C., 2017. Improvements in gold ore cyanidation by
pre-oxidation with hydrogen peroxide. Miner. Eng. 108, 67e70.
Rahimpour, M.R., Asgari, A., 2008. Modeling and simulation of ammonia removal
from purge gases of ammonia plants using a catalytic Pd-Ag membrane reactor.
J. Hazard Mater. 153 (1e2), 557e565.
Roberto, F.F., 2016. Commercial heap biooxidation of refractory gold ores e revis-
iting Newmont's successful deployment at Carlin. Miner. Eng. 106, 2e6.
Senanayake, G., 2004. Analysis of reaction kinetics, speciation and mechanism of
gold leaching and thiosulfate oxidation by ammoniacal copper(II) solutions.
Hydrometallurgy 75 (1e4), 55e75.
Senanayake, G., 2005. Gold leaching by thiosulphate solutions: a critical review on
copper(II)ethiosulphateeoxygen interactions. Miner. Eng. 18 (10), 995e1009.
Stenson, J., 2006. Disaster management as a tool for sustainable development: a
case study of cyanide leaching in the gold mining industry. J. Clean. Prod. 14 (3),
230e233.

squez-Lo
Vela
pez, P.C., Veiga, M.M., Klein, B., Shandro, J.A., Hall, K., 2011. Cyanidation
of mercury-rich tailings in artisanal and small-scale gold mining: identifying
strategies to manage environmental risks in Southern Ecuador. J. Clean. Prod. 19
(9e10), 1125e1133.
Wichink Kruit, R.J., Aben, J., de Vries, W., Sauter, F., van der Swaluw, E., van
Zanten, M.C., van Pul, W.A.J., 2017. Modelling trends in ammonia in The
Netherlands over the period 1990e2014. Atmos. Environ. 154, 20e30.
Xu, Y., Schoonen, M.A.A., 1995. The stability of thiosulfate in the presence of pyrite
in low-temperature aqueous solutions. Geochem. Cosmochim. Acta 59 (22),
4605e4622.
Yang, Y.-b., Cui, L.-n., Li, X.-s., Li, Q., Jiang, T., Ge, J., 2013. Novel technology on
preparation of double-layered pellets for sulfur and arsenic-bearing gold con-
centrates. J. Cent. South. Univ. 20 (11), 2967e2973.
Yu, H., Zi, F., Hu, X., Zhong, J., Nie, Y., Xiang, P., 2014. The cop-
pereethanediamineethiosulphate leaching of gold ore containing limonite
with cetyltrimethyl ammonium bromide as the synergist. Hydrometallurgy
150, 178e183.
Zhang, D.-c., Xiao, Q.-k., Liu, W.-f., Chen, L., Yang, T.-z., Liu, Y.-n., 2016. Acid leaching
decarbonization and following pressure oxidation of carbonic refractory gold
ore. J. Cent. South. Univ. 23 (7), 1584e1590.